Analytical and Interpretive Practices in Design
and New Product Development
Evidence from the Automobile Industry
by
Kamal M. Malek
S.B., Mechanical Engineering
Massachusetts Institute of Technology, 1982
S.M., Mechanical Engineering
Massachusetts Institute of Technology, 1985
Submitted to the Department of Mechanical Engineering
in Partial Fulfillment of the Requirements for the Degree of
DOCTOR OF PHILOSOPHY
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
June 2001
ARCHIVES
© 2001 Kamal M. Malek. All rights reserved.
The author hereby grants to MIT permission to reproduce
and to distribute publicly paper and electronic
copies of this thesis document in whole or in part.
Signature of Author:
--
wv
-
A
-
-
/MIT Deppinelt ofanical
Engineering
March 2001
/
Certified by:
Richard K. Lester
Professor of Nuclear Engineering
. Thesis Supervisor
Accepted by:
Ain A. Sonin
Chairman, Department Committee on Graduate Students
2
Analytical and Interpretive Practices in Design
and New Product Development
Evidence from the Automobile Industry
by
Kamal M. Malek
Submitted to the Department of Mechanical Engineering in March 2001 in partial
fulfillment of the requirements for the degree of Doctor Of Philosophy
Abstract
Product design and development have been studied from both positivist and interpretivist
paradigms. From the positivist perspective, design and product development are seen as
technical transformation or production processes, which take customer requirements and
existing technological possibilities as inputs and produce an objectively optimal product,
one that is not influenced by the designer's preferences or biases. The result of this
research is a focus on measuring the voice of the customer with "high fidelity", and on
streamlining and optimizing this production process. From the intrepretivist perspective,
product design and development are seen as relatively open-ended discursive processes,
to which human participants from different backgrounds bring their unique worldviews
and prejudices. Models of these processes are seen as metaphors intended to help people
come to understanding by shedding light on and thus bridging the different worldviews,
not as mathematical constructs to be optimized. In real life, empirical evidence shows
that practitioners rely on a number of approaches that do not fit easily into one or the
other of these paradigms. As a result, many analytical models and methodologies need to
be modified to make them useful in real-world applications and, coniverscly,empirical
research that accurately captures the richness and complexity of the design and
development process fits uneasily in these traditional paradigms which researchers feel
compelled to use.
This dissertation addresses this shortcoming by developing a vocabulary for describing
product design and development practices, which bridges the divide between the strictly
positivist and strictly interpretivist views. The research approach used is one of theory
building from case studies. The industry chosen for the case studies is the automobile
industry. The thesis reports on three study sites. The first is an American manufacturer
based in Detroit, known for its innovative product designs and its pioneering reliance on
dedicated platform development teams. The second is the American design subsidiary of
a Japanese manufacturer, one of the first to set up such a design operation in US. The
3
third site is the Japanese design and development organization of the same manufacturer,
based in a technical center outside of Tokyo.
The theoretical framework presented in this dissertation, which co-evolved with the
above case studies, takes the form of a taxonomy of product development practices. This
taxonomy draws upon concepts from linguistics and the philosophy of language. In a first
step, the distinction within linguistics between the structural sub fields (e.g., syntax and
semantics) and the functional sub field of Pragmatics is used to sharpen the difference
between analytical/structural practices on the one hand, and interpretive practices on the
other. In a second step. two views of interpretation, one grounded in linguistics
(Pragmatics. specifically). the other in the philosophical hermeneutics of Heidegger and
Gadamer are used to expand the interpretive category into two, referred to as pragmatic
interpretation and hermeneutic interpretation, respectively.
Each of the three case studies provides a good illustration of a product development
organization that relies predominantly on one of the types of practices and approaches
captured by the taxonomy. The findings suggest a number of recommendations for design
and product development managers and practitioners, as well as several directions for
future research.
Thesis Committee:
Warren P. Seering (Chairman)
Professor of Mechanical Engineering, MIT
Richard K. Lester (Thesis Supervisor)
Professor of Nuclear Engineering. MIT; Director. Industrial Performance Center
Michael J. Piore
David W. Skinner Professor of Political Economy. MIT
4
In memory of my father,
Moussa A. Malek
S
6
Acknowledgements
I wish to thank the members of my thesis committee-Warren
Seering, Richard Lester,
and Michael Piore-for their wise advice, effective coaching, and continuous
encouragement throughout this effort. Their steadfast support enabled me to complete
this work despite the exigencies of the last few years. I am especially thankful to Richard
and Michael for initially inviting me on this voyage of intellectual discovery, and for
introducing me to a world of thought and knowledge of which I had not been cognizant.
This research would not have been possible without the participation of a number of
respondents-designers, managers, and executives-at the companies discussed in this
thesis, and at the companies studied in the course of a larger project at the MIT Industrial
Performance Center on the organization of design and product development. By
generously giving of their time, and by sharing their experiences and insights, they
provided the empirical material in which this research is grounded. This thesis would also
not have been possible without the generous financial support provided me by the MIT
Industrial Performance Center.
I am grateful to my friends at MIT, and especially at the IPC, for cheering me on
throughout this endeavor and for enriching my experience there.
Finally, I would like to thank my friends and colleagues at NewcoGen and Affinnova-in
particular Noubar Afeyan-for affording me the opportunity to complete this thesis while
helping launch a new enterprise, thus giving me a chance to immediately put the
knowledge gained during this research to good use.
7
8
Table of Contents
Abstract ...............................................................................................................
3
Acknowledgements ............................................................................................
7
Table of Contents ..................................
9
..............................................................
List of Figures ...................................................................................................
13
List of Tables .....................................................................................................
14
Chapter I: Introduction .....................................................................................
15
Background ................................................................................................................. 15
Example 1: A minivan development project......................................................... 16
17
........................................
Example 2: Who's still buying sports cars? .
A Paradox in the Literature? .......................................................................................
18
23
.........................................
Different Paradigms ........................................
23
..............................................................................................
The design literature
The product development literature ...................................................................... 25
The Opportunity .......................................................................................................... 26
28
Research questions .........................................
Approach ............................................................................................................... 29
30
A Preview of the Dissertation .........................................
Chapter II: Research Approach ........................................
33
The Larger Project....................................................................................................... 34
36
Theory Building and Case Studies .........................................
Theoretical Framework .........................................
38
39
Empirical Research .........................................
The choice of industry ..................................................................................... 39
The cases ................................................................................
43
Chapter III: Theoretical Framework .................................................................
47
50
Background I: Linguistics .........................................
Structural linguistics.............................................................................................. 50
Syntax ..............................................................................................................
Semantics
.........................................
51
54
Functional linguistics ............................................................................................ 56
.........................................
Pragmatics
57
........................................................ 61
Pragmatics, more generally .........
70
Background II: Hermeneutics .........................................
The three meanings of hermeneutics..................................................................... 71
9
Interpreting as 'saying' ................................................
71
Interpreting as 'explaining' ........................................
72
Interpretation as 'translation' ........................................
........
74
The evolution of the field ........................................
..................................... 74
Key elements of hermeneutic interpretation.......................................................... 77
Contextual factors in hermeneutic interpretation ............................................78
Some important aspects of hermeneutic interpretation ...................................85
The Typology ..............................................................................................................90
Analytical or structural practices ...........................................................................93
Examples of analytical processes in design.....................................................95
Interpretive practices: pragmatic and hermeneutic................................................ 96
The differences between pragmatic and hermeneutic practices ...................... 98
Examples of interpretive practices in design.
.................................
100
Structural, pragmatic, and hermeneutic practices: a summary ............................ 102
The typology at work in PD ................................................................................
105
Stage gate systems......................................................................................... 106
DSM (Design Structure Matrix) ....................................................................
Chapter IV: Case Study I................................................................................
The NS Minivan Team ...................................................
His master's voice: VoC, QFD, and all that ...................................................
"Gut feel" and glory ............................................................................................
112
121
121
123
124
Innovation and the conflict between VoC and "gut feel".................................. 1..25
The answer to the dilemma: interpretation ...................................................
What makes interpretation work in this case? The team .....................................
126
128
Conviction, trust, and innovation .......................................
.................................
132
The Design Office ...................................................
138
Creativity v. practical considerations ...................................................
140
Pacifica: creativity though isolation ........................................
............................
140
Design decisions: conviction, responsibility, and what the boss wants .............. 142
Why have a separate DO? ...................................................................................
1. Developing and exploring new ideas .......................................
2. The DO's integrative role .................................................
147
147
150
Supplier Relations at Chysler: Analytical v. Interpretive Integration...................... 154
Chapter V: Case Study II................................................................................
161
History of NDI .............................................................................................
. . . . . . . .......
162
NDI as a Creative Entity ..................................
165
Creativity: a marriage between the rational and the intuitive ............................ 165
"Creative abrasion" ...............................................
167
Management strategies and leadership at NDI .................................................... 167
"Stepping back from the canvas" ........................................
.......
168
"Cheating" ..................................................................................................... 169
"Blurring the boundaries"............................................................................. 170
Strategies and practices that encourage open-ended thinking ............................. 173
Strategies that encourage interpretation or interpretive thinking .................. 177
10
The Work of a Design Studio: The Altima Project ................................................... 180
The Design Context Laboratory ....................................................
187
The history of the DCL .............
.......................................
187
The DCL and internal integration ....................................................
188
An example of the DCL's internal translation and mediation roles.............. 189
Other examples of internal integration.......................................................... 192
The DCL and external integration.......................................................................
194
DCL's approach to consumer clinics: the case of the Maxima ..................... 196
A second example of open-ended consumer interaction: the Altima............ 202
A summary of the DCL's philosophy and approaches........................................ 203
The Color Studio ....................................................
206
Examples of internal integration ........................................
............
208
The Nissan global color meeting................................................................... 208
Applying colors and textures to forms that don't exist ................................. 212
Examples of external integration.........................................................................217
"Market segmentation" by bag of"chips": Understanding sports car buyers217
Drive-by shooting: A photo essay of random cars and their drivers ............. 225
Watching Oprah to stay in touch with the public .......................................... 226
Chapter VI: Case Study III ..............................................................................
229
Product Planning Department #2 ........................................
...............
Exterior Design Department #2.......................................................
229
232
Product Experiment Department #2 .......................................................
236
Chapter VII: Summary and Conclusions .....................................................
Findings .....................................................................................................................
241
241
Theory development............................................................................................ 241
Empirical findings .......................................................
243
Limitations .......................................................
246
Implications ............................................................................................................... 247
For researchers .......................................................
247
For PD managers and practitioners ........................................ ...............
248
Suggestions for Future Research .......................................................
Bibliography .................................................................................
II
252
255
12
List of Figures
Figure III.1: Schematic of different levels of context (based on description in Akmajian et
al., 1995) ...........................................................
67
Figure 111.2:Stage-gate system model (from Cooper, 1990) .......................................... 107
Figure 111.3:The relationships between design tasks (from Eppinger et al., 1994) ....... 114
Figure III.4: The original design structure matrix (from Eppinger et al., 1994) .....
1....
115
Figure 111.5:The DSM after re-sequencing of tasks and partitioning (ibid.) ............... 115
13
List of Tables
Table 1I. : The case studies sites ...........................................................................
45
Table 111.2:The different kinds and levels of context (based on Dascal, 1989) ............... 68
Table 111.3:A comparison of an analytical and an interpretive model of design: Simon's
rational problem solving model and Sch6n's reflection-in-action model
(Reproduced from Dorst and Dijkhuis, 1996) ..................................................... 101
Table III.4: Analytical v. interpretive approaches: a summary ....................................... 103
Table 111.5:Analytical, pragmatic and hermeneutic practices: key features ................... 104
14
Chapter I: Introduction
Background
This dissertation is a study of certain practices in design and new product development,
activities in which most industrial companies engage on an ongoing basis, and which are
credited with a significant part of the success of many of them. In recent years, various
studies have discussed the importance of the design and development phase of a given
product. Studies carried out by Ford and Xerox have underscored the impact of the
design and development phase on the cost and quality of a product, and on the time it
takes to bring that product to market, all of which are critical factors in its ultimate
performance in the marketplace (Ullman, 1992). The last decade has seen a dramatic
increase in the number of studies, academic papers, articles and books published on the
subject (Cooper & Kleinschmidt, 1986; Takeuchi & Nonaka, 1986; Hollins & Pugh,
1990; Clark & Fujimoto, 1991; Dougherty, 1992a; Wheelwright & Clark, 1992; Bowen,
Clark, & al., 1994; Deschamps & Nayak, 1995). Even the popular business press has
recognized the importance of product design as a source of competitive advantage. In
1988, an issue of BusinessWeek was dedicated to the subject and, starting in 1991, that
periodical teamed up with the American Association of Industrial Designers to sponsor
their yearly Design Excellence Awards and to publish the results in a special issue
(Nussbaum, 1993; 1997).1These awards recognize the best product designs in a number
of industrial categories, from automobiles to computers to kitchen utensils.
Designing and developing a new product is a complex activity, involving a wide range of
tasks, including sketching, drafting, sculpting clay models, performing complex
calculations and computer simulations, analyzing competitor's products, deciding on
See BusinessWeek issue dated April 11, 1988. Also see the issues on the Annual Design Awards, such as
those dated June 17, 1991 and June 2, 1997.
15
product features, cost, and market positioning, to name a few. Perhaps one of the most
important activities undertaken by product designers is finding out what the prospective
users of their creation might be looking for in the new product, and learning about their
needs, desires, and preferences. The importance of this activity is reflected in the findings
and recommendations contained in the body of research of the last decade. For example,
in their article on the importance of "product integrity", Clark and Fujimoto define the
external dimension of product integrity as "the consistency between a product's
performance and customers' expectations" (Clark & Fujimoto, 1990, p. 108). And in his
monograph on product development, Clausing warns against any "disregard for [the]
voice of the customer", and states: "High fidelity to the voice of the customer is a key
success factor." (Clausing, 1993, p. 12). The following examples describe two instances
of product developers involved in interactions with their prospective customers, and
suggest how complex these interactions are in reality.
Example 1: A minivan development project
This example concerns one particular decision in the development of the previous version
of a popular minivan, the 1996-2000 Chrysler Minivan; specifically, the decision to give
the minivan a fourth door, that is, a sliding door on the driver's side of the vehicle.
Previously, no minivan had offered such an option. The feedback that the designers were
getting from their consumer research suggested that the fourth door would not be well
received at all, that it would be a "loser". Not only did the clinic participants feel that it
was not needed, many of them were vehemently opposed to it on safety grounds, as they
pictured their children jumping out of the minivan in the path of oncoming vehicles. The
designers, on the other hand, believed that it would be a very useful and practical addition
to the vehicle. They decided to go against their market research data and to include the
fourth door as an option. After the vehicle was released, the fourth door quickly became a
very popular option. Indeed, it changed the dynamics in that market segment, as
competitors scrambled to redesign their minivans to offer a fourth door option too.
Analysts were soon explaining the loss of market share by a competitor's highly regarded
minivan as a direct result of the unavailability of such an option (Phillips, 1998).
16
It should be noted that this does not appear to have been an isolated case at Chrysler. In a
recent book, the then Chrysler vice-chairman, Robert Lutz, discusses some of the "laws"
that made the company successful (Lutz, 1998). Among them: "The customer is not
always right."
Example2: Who's still buying sports cars?
The second example concerns another automobile development project, this one
involving a new sports car. In the last decade, the market share held by sports cars in the
US has declined steadily, to the point where one manufacturer after another abandoned
the segment. In the process, some very well known and highly regarded models
disappeared, such as Mazda's RX-7, Nissan's Z car, and Toyota's Supra.2 Tasked with
developing a new sports car that would help reverse that trend, some designers at Nissan
Design International chose an unusual way to learn about their potential customers.
Instead of sending a questionnaire to current sports car owners or holding a focus group
(Krueger, 1994), they decided to go around a number of parking lots in different parts of
the country, choosing among the parked cars those sporty cars they deemed "interesting",
and leaving an envelope of "stuff' on the windshield, along with a letter of explanation
and token compensation. These bags contained small pictures of a number of items
ranging from chairs of different styles, to a miniature plastic hamburger and Aunt Jemima
pancake syrup bottle, to apparel items and accessories such as sunglasses. The owners of
these cars were asked to sort the items in the bag according to whether they liked,
disliked, or simply did not care about them. Interestingly, the designers were aware of the
biases their approach entailed, but were unconcerned about them. More importantly,
when asked to explain their choice of the items included in the bag, their answer was that
they simply liked these items or thought they were "interesting". Similarly, they did not
have any specific plan for analyzing the answers or for using the resulting information in
their design work. However, they expected that by going through the material returned by
the respondents they would understand them better, though they were not able to
2
This shift in consumer choice has been explained as resulting from the increased popularity of sport-
utility vehicles and the changing demographics with the aging of the baby boomers, among other factors.
17
articulate that understanding beyond saying that a particular choice "makes sense" in
light of the other choices made by that respondent.
Comparing his designers to the actor Anthony Hopkins who played the role of a
cannibalistic murderer in the movie "The Silence of the Lambs", the president of Nissan
Design International explained that approach to consumer research thus:
We'renot designingfor the customer.Weare methodactors. Weget into
the framework, into the mindset [of the customer] to the extent that's
possible.
The first of these examples shows that faithfully following the voice of the customer is
not necessarily the only way to a successful product, even where a well-defined design
decision is con erned. The Minivan team did indeed "listen to the voice of the customer;"
they simply chose not to obey it mechanically. This example also shows that focusing
purely on meeting customers' previously formulated expectations would not leave much
room for surprising them with novel and innovative ideas.
The second example indicates that many designers do not see their work as simply
soliciting the voice of the customer and following it with "high fidelity". The information
they seek about the customer is not specific. or it may be exploratory in nature and not
targeted to particular issues or concerns that are well defined by the design team. The
example also suggests that it is not obvious how that information is ultimately used in the
design process.
A Paradox in the Literature?
The examples in the previous section suggest that there are many ways in which
designers and product developers learn about their customers and come to understand
them and their preferences. The examples also show that there is more than one way in
which designers use (or not, as the case may be) what they learn about their customers, to
the point where, on the face of it, the admonition to "listen to the voice of the customer"
means little. Beyond this, however, the examples also reflect a deeper paradox in the
product development literature, which I outline below.
18
In their review of the management literature on product development, Brown and
Eisenhardt identified three distinct streams of research (Brown & Eisenhardt, 1995).3 The
first stream, product development as rational plan (or the "rational stream"), is not
particularly interesting for this discussion, as the relevant research tends to be largely
atheoretical. The second stream, the "communication stream" (or product development as
a communication web), is more interesting. Its basic premise is that greater
communication among development team members, and between them and outsiders,
leads to greater success.4 Yet within this stream we find widely disparate notions of what
communication is and how to describe and quantify it. At one end of the spectrum, we
find the early work by Allen (Allen, 1971), the pioneering work from which this
particular stream evolved, according to Brown and Eisenhardt's classification. We also
have the closely related work of Katz and Tushman (Katz & Tushman, 1981). In both of
these studies, the level of communication is measured by asking the subjects (engineers,
scientists, and managers) to keep track of, and report to the researchers, the number of
communications they had over a given period of time. The content of the particuiar
communication is not considered, nor its importance for the project (though respondents
are asked not to report insignificant chats.) In fact, since every respondent was asked to
3 Following Adler, product development in the management literature is considered a subset of the
literature on innovation (Adler, 1989). The latter is divided into two general areas of research. The first is
innovation proper which is concerned with differences in patterns of innovation across sectors and
geographical areas, with an economics orientation. The second, product development, is concerned with
how specific new products are developed within particular organizations (Brown et al., 1995).
4 The
importance of communication
and information in new product development stems from one of the
defining characteristics of that activity, namely, the uncertainty that pervades it. As they seek to create
and sell what does not yet exist, product developers face uncertainties from various sources. These
include, among others, uncertainties about the market they are targeting, and the preferences, wants, and
needs of their potential customers; uncertainties about changes in that market and these preferences
during the gestation period needed to develop and introduce the product; uncertainties concerning the
availability of the technologies they plan on using; and, of course, uncertainties about their competitors
response and the products they will be introducing around the same time. Information plays an important
role in helping developers deal with these uncertainties, based on a generally accepted relationship
between information and uncertainty. That relationship holds that the likelihood of success of a design
or, more generally, an uncertain endeavor or decision, is proportional to the information that is available
to those undertaking that design or making that decision (Suh, 1990). More precisely, the likelihood of
failure increases as the information gap. that is, the difference between the information needed for the
design, and the information available, increases. Hence the focus in the literature on getting information
from the customer, from the suppliers, and from the different technical specialists, and on coordinating
the exchange of information between the various people and decision-makers involved in the
development project. An important aspect of this relationship between information and risk reduction is
that information is seen as an undifferentiated entity, with more of it, in whatever form, always being
better. The association of information with uncertainty and uncertainty reduction is a basic tenet of
communication or information theory (Pierce, 1980).
19
track his communications on one particular, randomly selected day of the week, it would
not have been possible to say anything about how a particular communication
fit in
within, or how it impacted, any ongoing problem or issue that the team was dealing wiil.
The other end of the spectrum in this communication stream is represented by the work
by Ancona and Caldwell (Ancona & Caldwell, 1992), and by Dougherty (Dougherty,
1992a). These researchers actually examine the content of these communications to
develop a deeper understanding of how and why the different communication and
integration mechanisms work. Dougherty's work, for example, is concerned with crossfunctional integration in view of the different worldviews or thought worlds and systems
of meaning of the different functional departments involved in the development project.
She is concerned with the way in which people from different functional areas focus
differentially on the various aspects of the product, how and why they interpret a given
piece of information differently, and how their different perspectives ultimately come
together in a successful project. Her research is not only concerned with the content of
acts of communication between people, but also with a number of contextual factors
within which they take place.
One could argue that these differences in the way researchers conceive of communication
in product development is merely a reflection of the particular classification scheme used
by Brown and Eisenhardt to organize the literature. I believe that the problem goes
beyond that, however, as it is not uncommon to find discordant notions of
communication showing up in the very same paper. A paper by Moenaert and Souder in
which they develop an information transfer model and use it to discuss different
mechanisms for integrating the marketing and R&D functions in new product
development projects, is one such example (Moenaert & Souder, 1990). The model of
communication they present follows the information transfer practice of equating
information with uncertainty reduction (Pierce, 1980; Suh, 1990). According to that view,
information is pre-existent, in the sense that someone out there knows or has the
information that is needed by someone else; as the authors put it: ... every individual is
viewed as a pool of knowledge." The organizational integration challenge then is one of
getting the relevant parties to exchange information in a timely and efficient manner.
20
From that perspective, all integration mechanisms end up being reduced to one metric,
namely, their effect on the frequency and likelihood of information transfer. Integrative
mechanisms ranging from the technical, such as task specification and planning, to the
more interpretive boundary spanning mechanisms such as job rotations and the use of
multidisciplinary teams, and even cultural change programs intended to improve
harmony and trust within the organization, all become quasi-interchangeable. The authors
simply assert that:
In technological innovation, the major merits of each of'the above
discussed integration mechanisms consists qf the improvements in the
interfunctionalinfbrmation ransfer that thesemechanismsbringabout.
(Page 97.)
And yet, in the same paragraph, the authors describe the role of these integration
mechanisms in creating a social network in which people from the different functional
groups become connected through a variety of relationships:
Through these interpersonal linkages, [a]ffects, influence, information and
goods and services will be exchanged.5 (Page 97.)
and
Infirmation transfergeneratesa mutual understandingof each other 's
roles and helps determine what is expected of everybody. (Page 99.)
It is difficult to imagine how to reconcile the standard analytical information-transfer
model of communication with the rich and complex social interactions from which such
highly personal and emotional reactions as affect, mutual understanding, and influence
might evolve.
This inconsistency in the use of key underlying concepts is not limited to the
communication stream of the product development literature. It is also found in the third
and last stream in Brown and Eisenhardt's taxonomy, which they refer to as the "problem
solving stream" (or design as disciplined problem solving.) One of the best-known
contributions in this category is Clark and Fujimoto's compelling study of product
5
"Effects" in the original. I believe that to be a typo.
21
development performance in the automobile industry (Clark et al., 1991). A key finding
of this study is the critical role the "heavyweight product manager" plays in the success
of Japanese automobile development projects. In the second chapter of their book, the
authors explain how their research is based on an information-processing perspective of
product development, as opposed to the more traditional material-processing
perspective
(pp. 20-22).6 Later on, however, they prescribe the following qualifications and roles for
the heavyweight project manager: he possesses "market imagination and the ability to
forecast (...) based on ambiguous and equivocal clues;" he knows how and when to
"initiate conflict" in order better to direct the evolution of the product design; and he is
skilled at "envisioning the desires of the customer and interpreting them for the other
members of the development team" (p. 343). From an analytical perspective, it is not
clear what kind of information processing. be it a mathematical transformation or an
algorithmic procedure, would be involved in initiating conflict between members of a
development team, or in envisioning customer desires.
How might a reader, in particular a product development practitioner, react to these
inconsistencies? The pragmatic reader, recognizing the limitations of the theoretical
model, might simply discount the problem and move on to the more important and richer
empirical findings. A more literal reader might take the claimed theoretical underpinnings
at face value, see them as foundational, and view the empirical findings and
recommendations through this information-transfer or information-processing lens. Such
a reader would likely end up discounting some of the more insightful and significant
findings presented by the authors, which do not readily fit this theoretical model. Still
another type of reader might be concerned about the inconsistency between the theory
and the empirical findings and, as a result, lose confidence in the otherwise excellent
recommendations given. In any case, this mismatch between research findings and
recommendations on one hand, and the theoretical models that are presumed to underlie
them on the other, gives the former a rather ad hoc characier. and makes it more difficult
for the practitioner to assess their broader relevance and applicability.
6
It should be noted that Clark and Fujimoto themselves link their work to the communication stream of the
product development literature. They trace the information paradigm they use to Allen's and Marquis'
work (Allen, 1977; Marquis, 1982) (Ch.2, note 2.)
22
Different Paradigms
The inconsistency in the product development literature pointed out in the previous
section reflects a deeper issue in the study of design and product development, namely,
what the proper paradigm for such study ought to be (Kuhn, 1970). 7 The two major
paradigms of relevance here, positivism and interpretivism, have already been hinted at in
the previous section. In the case of communication for example, the former gives us the
notion of communication as information transfer and formal, codifiable information
processing, with a resulting focus on channel capacity, frequency of transmission and
other similar analytical measures (Pierce, 1980). The latter leads us to view
communication as an idiographic social exchange between complex human beings with
distinct worldviews, concerns, interests, and affects. Different researchers in design and
product development tend to pursue their work from within one of these two paradigms.
What follows is a partial review of the literature from the perspective of these two
paradigms. In the first subsection, I look at the design literature. In the second subsection,
I return to the product development literature I began discussing in the previous section.
The design literature
In research on design in engineering and architecture, the logical-positivist paradigm has
been dominant, starting with Simon's admonition to researchers to develop a 'Design
Science' (Simon, 1981). Research that fits within this paradigm includes work on design
methods by Simon himself, Alexander's early work (Alexander, 1964), and others'
(Jones, 1981; Rowe, 1987). Other researchers working within this paradigm have focused
on developing process models for design engineering (Pahl & Beitz, 1984; French, 1985;
Pugh, 1991; Ullman, 1992). Still others have developed theories of design (Hubka &
Eder, 1988; Suh, 1990).
Within the positivist paradigm, design activities and processes follow logical, quasiformal models, which makes them particularly suited to codification in computer
programs. As a result, much of the work within this paradigm has involved computer
(artificial intelligence) systems. Examples include: the work on generative shape
7 Paradigm is used here in its social sense, as "the entire constellation
on shared by the members of a given community" (Kuhn, 1970).
23
of beliefs, values, techniques, and so
grammars or 'algebras' (Stiny & Mitchell, 1978; Stiny, 1980; Mitchell, 1990; Wallace,
1991a); the work on capturing and codifying existing expert knowledge in rule-based and
expert systems (Gero, 1989; Coyne, Rosenman, Radford, Balachandran, & Gero, 1990;
Navinchandra, 1991; Tong & Sriram, 1992a, b); research that takes the connectionist
approach to AI (McClelland & Rumelhart, 1988), such as (Coyne, Newton, & Sudweeks,
1993b); and research that uses genetic and evolutionary computational algorithms to
perform design exploration and optimization (Coyne et al., 1993b; Gero & Kazakov,
1996; Funes & Pollack, 1997).
Within the same positivist paradigm, some researchers have focused on the human in the
design process, and have used protocol studies to obtain detailed descriptions of the
designers' activities (Ericsson & Simon, 1993). These studies involve observing. timing,
and coding the activities and statements of designers working under controlled
conditions, again for the purpose of codifying and developing analytical models of what
they do (Akin, 1978; Akin & Lin, 1996; Dorst & Dijkhuis, 1996).
The interpretivist paradigm has been less prominently represented in the engineering and
architectural design literature. SchSn, in response to what he saw as shortcomings in the
design methods approach advocated by Simon. showed the designer as someone involved
in a reflective, interpretive practice, not merely a process of mechanistic information
processing (Sch6n, 1983). Dorst has applied Sch6n's model of design practice in a
comparative protocol study (although still using the positivist techniques of protocol
analysis) (Dorst et al., 1996). Bucciarelli studied a group of engineers and described their
work and interactions (Bucciarelli, 1994).
With the increasing importance of cooperative work and collaborative computing, many
researchers have turned their focus on these aspects of design practice and their
implications for the development of tools and environments for collaborative design
(Snodgrass & Coyne. 1992; Fischer. Nakakoji, Ostwald. Stahl. & Sumner. 1993: Stahl.
1993; Coyne, 1995, 1998; Gallemore. 1998).
Researchers have also looked at industrial design from an interpretivist perspective
(Krippendorff & Butter, 1989; Jonas. 1993). Krippendorff has described the role of the
24
various contexts-the context of genesis and the context of use for example-that affect
the design of a product (Krippendorff, 1989).
Theproduct developmentliterature
Turning to the management literature on product development, we find an important
body of work that falls within the positivist paradigm. This includes the research
identified in Brown and Eisenhardt's review of the literature as falling under the "rational
stream." This stream builds on the work of Myers and Marquis and the British SAPPHO
studies (Myers & Marquis, 1969; Rothwell, 1972; Rothwell et al., 1974). More recent
work in that stream includes work by Cooper and Kleinschmidt, and by Zirger and
Maidique (Cooper & Kleinschmidt, 1987; Zirger & Maidique, 1990). I consider this
research as positivist in nature more as a result of the methods used in the studies-
statistical analysis-than of the findings or conclusions that result. Since this research
tends to be atheoretical, there are no underlying models or metaphors of the product
development process as a whole, or of communication or integration mechanisms.
Another body of work, not discussed in Brown and Eisenhardt's review, stems from the
introduction of concepts and methodologies from engineering to address organizational
issues in product development. These include: the use of system dynamics models of
certain processes in product development (Kim, 1993); the use of stage-gate systems to
organize the development process (Cooper, 1990);8 the use of linear algebra models to
represent the sequencing of tasks in a project (Eppinger, 1991; Eppinger, Whitney, Smith,
& Gebala, 1994); and techniques used to plan for product modularity and part
commonality (Robertson & Ulrich, 1998). I would also place in this category the
techniques developed to map market research data into product features, such as 'Quality
Function Deployment' (the 'House of Quality') (Hauser & Clausing, 1988; Clausing,
1993), and Burchill's 'Concept Engineering' process model (Burchill, 1993; Burchill &
Fine, 1997).
Under the interpretivist paradigm, we find: the work by Dougherty, some of which was
previously mentioned (Dougherty, 1990, 1992a, b; Dougherty & Heller, 1994); similarly,
25
the work by Ancona and Caldwell (Ancona & Caldwell, 1990; Ancona et al., 1992); and
the work by Carlile on the use of 'boundary objects' by product development teams
(Carlile, 1997).
One field of research related to product development in which both paradigms have been
prominent is the field of consumer research. The methods of interpretivist research in that
field are discussed a number of authors, among them: Hirschman (Hirschman, 1986);
Hudson and Ozanne (Hudson & Ozanne, 1988); Thompson et al. (Thompson, Locander,
& Pollio, 1989; Thompson, Pollio, & Locander, 1994); Arnold and Fischer (Arnold &
Fischer, 1994); and Holbrook, who provides an example of the use of an interpretivist
approach to understand movie going (Holbrook & Grayson, 1986).
The Opportunity
In general, the literature on design and product development is reminiscent of Kuhn's
'rival paradigms' in science: "incommensurate domains of discourse largely ignorant of
each other..." (Kuhn, 1970; Coyne, 1995 p.206). Researchers within each paradigm have
their own models and metaphors for conceiving of and studying design and product
development. We also have debates over the proper paradigm for such studies (Snodgrass
& Coyne, 1990; Coyne & Snodgrass, 1991, 1993a; Piore, Lester, Kofman, & Malek,
1994), reflecting similar debates in the broader area of management research (Morgan &
Smircich, 1980; Beyer, 1992; Donaldson, 1992).
These are not idle debates among academics. Researchers engaged in their work stop
seeing these models and metaphors for what they are, and start applying and propounding
them literally, through a sort of "inversion of reality", or what Meyer refers to as the
"suspension of metaphoric thinking," with the resulting transformation of "theories into
myths" (Hesse. 1980; Meyer, 1984).9 Yet these researchers are not philosophers or
mathematicians arguing and debating within a closed community, removed from practical
8
The stage-gate model can be thought of as a mapping of Pahl & Beitz' engineering design process model
from the product domain to the domain of the organization developing it (Pahl et al., 1984).
9 For a discussion on how metaphors shape our thinking in general, see Lakoffand
Johnson (Lakoff&
Johnson, 1980). For a more specific example relating to communication, see Reddy (Reddy, 1979) and
Axley (Axley, 1984).
26
concerns. Their conceptions of design and product development shape the prescriptions
they propound to practitioners, and affect how the latter come to understand what they do
and how they should do it. Often, the result is a distortion of the design and development
process as practitioners seek to reify the concepts and models presented to them by the
researchers. This is particularly true in the case of the positivist paradigm, for two
reasons. One is the privileged position in which detached, 'objective', Cartesian
rationalism is held, especially among scientists, engineers, and many managers. The
second reason is that interpretivist prescriptions, such as the use of metaphor, are not as
easy to apply as the typical analytical methodology. Due to the idiographic nature of that
research, the creative burden of coming up with the appropriate metaphor and of
'operationalizing' it remains for the practitioner to bear (Dougherty, 1992b; Coyne,
1995).
For example, a positivist perspective can lead to an understanding of design and product
development as transformation process, with the needs and desires of the customer, along
with the technical possibilities available to the firm, as exogenous inputs to that process.
The role of the development organization ;s one of an ideal transformer carrying out a
purely technical, mechanical translation process. This understanding leads researchers to
develop prescriptions and methodologies for: a) obtaining that precious first input, the
'voice of the customer', with a high degree of "fidelity"; and b) for faithfully translating
that input without adding any distortion or spurious inputs of one's own (Burchill, 1993;
Clausing, 1993; Ulrich & Eppinger, 1995). In turn, this leads practitioners who follow
these prescriptions to "rely on research, not opinion" and to view designers who are too
attached to their product as a liability to the organization, not an advantage (Connelly.
1994; Automotive News, 1995; Child. 1996a).' 0
Certain researchers have presented "multi-paradigm" studies of design and product
development, in which models and metaphors stemming from the different paradigms are
proposed (Snodgrass et al., 1990, 1992; Jonas, 1993; Coyne, 1995; Dorst et al., 1996). In
'0 Reactions to such claims from product developers who see themselves as following a more pragmatic
approach tend to be equally extreme: they compare listening to the voice of the customer to "looking
through a rearview mirror," and they prefer to trust their intuition over any consumer clinic (Stout, 1995;
Child, 1996b; Yung, 1997; Lutz 1998).
27
general, the result is a juxtaposition of perspectives, suggesting that one can look at
design and product development through different lenses, each of which remains
embedded within its own worldview, but one of which is privileged. Effectively, the
practitioner is presented with an exclusive choice; the different perspectives remain
incommensurate, with no clear way for bridging the chasm between them. This is not
surprising, given the very definition of paradigm.
Aside from this situation of "rival paradigms"' in design and product development, we
also find many instances of what could be termed "confused paradigms", where the
vocabulary and concepts from different models of product development are mixed in a
muddled way. This situation, and the problems it can engender, were discussed in the
previous section, in the context of the use of the information transfer and information
processing models by Moanert and Souder and by Clark and Fujimoto (Moenaert et al.,
1990; Clark et al.. 1991). This situation is more likely to arise in empirically-based
product development research as in the cases above, than in more theory-driven work
where models are less likely to clash with the reality of the empirical observations. The
fact is that most practitioners, notwithstanding the claims and stated beliefs of some of
them, do not work within one paradigm, or according to one model or metaphor. They
deploy various models and metaphors, and engage in a range of activities, some of them
positivist, others interpretivist, as they face different situations. They pragmatically use
whatever tool or approach they feel they need in any given situation, as the two examples
at the beginning of this chapter showed.
Research questions
The question this dissertation poses, and attempts to answer, is the following: is it
possible to develop a vocabulary for describing the activities in which designers and
product developers engage in the course of their work which would bridge the gap
between the current positivist and the interpretivist metaphors? Is there a way to discuss
product development practices without placing them at one or the other end of the
positivist - interpretivist dichotomy? If we were to allow ourselves, for a moment, to
abuse the notion of incommensurability of paradigms, could we think of a taxonomy that
28
classifies practices along a continuum, fromnthe highly analytical or positivist, to the
highly interpretive? What form might such a taxonomy take?
From the preceding sections, the advantages of developing such a vocabulary or
taxonomy are clear. Instead of thinking of the various types of approaches as exclusive it
would be possible to think of them the way practitioners use them in real life. It would
become possible to ask questions about, and discuss. the relevance and applicability of
different types of practices to a given design or product development situation. By the
same token, such a typology would enable researchers and practitioners better to assess
the applicability of the various tools and methodologies to the problems they face. Some
of the questions one could pose: Are there different aspects of a given product for which
one type of approach is better suited than another? Are certain types of activities more
relevant at different phases of the development process? What demands do the different
types of approaches place on the practitioner and on the organization as a whole? Are
certain types of organizations better equipped to use one versus another type of
approach?
In his critical contribution to our understanding of the work of designers and other
professionals as interpretive practice, Sch6n described the professional alternating
between action and reflection on the situation or problem he or she faces (Sch6n. 1983).
With the proliferation of tools. methodologies, and best-practice recommendations
available to them I believe it is important for practitioners to be able to reflect not only
on the design situation at hand, but also on the tools and approaches they are using. A
vocabulary that can describe these approaches and practices as practitioners use them
would represent an important step towards extending Sch6n's concept of reflection-inaction from the task domain to the realm of practices.
Approach
The approach I use in this dissertation is one of theory building from empirical case
studies, as described by Eisenhardt (Eisenhardt, 1989). That approach draws on ideas
from grounded theory and case study research among others. It is particularly appropriate
here given that the aim of this research is not to test some hypothesis or to extend an
existing theory, but rather to develop new understanding. Theory building starts with the
29
research questions posed above, and involves repeated iterations between theory
formulation and data analysis, within cases and across cases. In this regard, the work to
be reported in the thesis fits within the context of a larger study of the organization of
product development (Piore et al., 1994).
For the theoretical framework, I draw mainly upon ideas from the fields of linguistics and
the philosophy of language to develop a taxonomy for classifying and describing the
various interactions and activities found in product development projects.
For the empirical research, I elected to study the automobile industry, mainly because of
the complexity of the product itself and the richness of its interface to the customer
(Clark et al., 1991). The design follows embedded multiple-case format, and the main
research instrument was qualitative, open-ended interviews.
A Preview of the Dissertation
The content of this thesis is organized in six chapters, including the present one. In this
introduction, I have introduced the research project. explaining the motivation behind it
and its objectives. I also presented an overview of the relevant literature and located this
work within it.
The next chapter, Chrpter II, is a description of the research approach and methods I used
in this work. In it, I discuss the reasons for selecting the particular industry and research
sites for the empirical part of this study. I also briefly present the larger research project
of which this dissertation is part.
Chapter III deals with the theoretical concepts from linguistics and the philosophy of
language upon which the theoretical framework of this research is based. This chapter is
intended to be a self-contained presentation, to make it accessible to product development
students and practitioners who may not have a background in these topics. The relevant
levels of linguistic analysis are presented and discussed. The topics of Pragmatics (from
linguistics) and Hermeneutics (from the philosophy of language) are of particular interest
in the research, and particular attention is paid to the differences between them. A
typology of design and product development approaches and practices that is based on
30
these concepts and that is illuminated by earlier empirical findings from the larger
research project is presented. The chapter closes with an illustrative application of this
typology to two bodies of research in product development methodology, the stage-gate
process and the design structure matrix.
Having laid out the theoretical framework for the dissertation, I report the empirical
findings from the various case studies in the chapters that follow. Each chapter
corresponds to one of the research sites. Chapter IV covers the cases conducted at the
Chrysler Corporation.
One case study deals with a new minivan development project.
(Example I at the beginning of this chapter offers a preview of that case.) Another study
focuses on Chrysler's Corporate Design Office.
Chapter V reports on the studies conducted at Nissan Design International (NDI) in San
Diego, a subsidiary of the Nissan Motor Company of Japan.' 2 The cases at NDI covered
the design of a midsize sedan design project (the Nissan Altima), the work of a market
research lab, and the work of the color and fabric studio (which provided the material for
example 2 above.) In addition, we also studied the NDI organization as a whole, its
inception, and the philosophy and organizing principles of its president.
Chapter VI reports the findings from the study conducted at Nissan's Japanese design
center, located at the Nissan Technical Center outside of Tokyo. Here again, the Altima
project was used as a focus for the interviews, but these covered a set of wide ranging
topics.
Finally, Chapter VII synthesizes and discusses some of the findings from the case studies.
It presents some conclusions and recommendations both for researchers as well as
managers and practitioners in product development. It also suggests a few avenues for
future research.
"Since the case studies were conducted at the Chrysler Corporation, the company has merged with
Daimler-Benz A.G. The combined company is now DaimlerChrysler A.G.
12
Recently, NDl's name was changed to Nissan Design America, or NDA.
31
32
Chapter II: Research Approach
This chapter discusses the research approach used in this dissertation. I start by
describing the larger research project on the organization of design and product
development, of which this dissertation constitutes one part. Following that, I briefly
discuss the theory building approach used throughout the larger project, and its
implications for the choice of research methods and sites. Finally, the choice of industry
and the selection of specific sites for this dissertation are discussed.
To begin, the objective of the dissertation as it was developed in the previous chapter is
briefly restated. That objective is to develop a taxonomy of design and product
development practices. My purpose in doing so is to bridge the gap between, on the one
hand, the rich and diverse range of activities which practitioners undertake in the course
of their work and. on the other hand, the models and metaphors commonly found in the
literature. The latter tend to be embedded within one of two paradigms-the positivist
and the interpretivist-or
to be couched in the limiting and sometimes misleading
languages of these paradigms. The activities I am interested in are those touched upon in
the previous chapter. These are integration activities that involve communication and
information transfer or processing on the one hand, and interpretation, understanding and
"sensemaking" on the other (Weick, 1995). More specifically, I am interested in these
integration activities as they take place across two distinct kinds of boundaries. The first
are those between the product development organization and its customers, actual or
potential; this I refer to as external integration. The second kind are those found within
the organization itself, between different functions or groups; this I refer to as internal
integration. This usage is somewhat idi,,,yncratic, as some authors use external
integration in the context of the interactions between the development team and the larger
organization and use internal integration to refer to what goes on within the team itself.
33
The Larger Project
The work presented in this dissertation is part of a larger project on the organization
of
design and product development. More generally, the objective of the project is to
develop insight into the processes and mechanisms of organizational integration under
conditions of radical uncertainty. Such situations cannot be modeled in terms of a set of
likely outcomes, each having a likelihood assigned to it based on some a priori
calculation or based on historical precedent (Knight. 1957). Rather. these situations are
too ill defined and equivocal, to the point where possible outcomes cannot be adequately
described at the outset. Since this is one of the defining characteristics of design and
product development projects. these were chosen as exemplars tor such situations and
served as the focus of research in that larger project.
3
A key aspect of the larger project is the theoretical framework it employs. Its basic
premise is that interpretation in general. and the field of hermeneutics in particular., offers
rich and meaningful metaphors for the processes that organizations use effectively to
address and manage situations of radical uncertainty. It identifies two types of integrative
approaches, the analytical and the interpretive. The analytical approach views design and
product development as structured problem solving, a process with well-defined stagesincluding a beginning and an end-and
governed by clearly defined means-ends
relationships. The interpretive approach on the other hand is more process oriented and
views integration as an ongoing process of coming to understanding, based on the
hermeneutic notions of the circle and the fusion of horizons (Palmer. 1969). (Please refer
to Chapter III.) This theoretical framework was presented in Piore et al. (Piore et al..
1994).
The larger project was centered on a set of cross-national. cross-industry empirical
studies. These have included a study of the cellular telephone equipment industry, the
fashion apparel industry. and the medical devices industry. These studies have covered
companies in the US, Japan, and Europe. In the first study. the two main sources of
uncertainty companies faced were the technological developments and the changing
'3 There are
many design and development situations where the outcome is not equivocal as described
above; these include routine design or parametric design. where the new artifact is the result of applying
slight modifications to an existing design.
34
industry and government and international standards. That study suggested a shift from
interpretive to analytical approaches connected with the lifecycle of a technology, in
particular that of cellular telephony. Early on in that lifecycle. as the possibilities afforded
by the technology and the nature of the market for it were still evolving and being
developed and defined, companies were more likely to rely on an interpretive approach to
product development and decision making in general. As the technology evolved and
matured, however, and the uncertainty associated with both the technology and the
market decreased, companies tended to shift to a more analytical approach to managing
the development process (Piore, Lester, & Malek, 1995). In that particular industry.
however, the technology does not appear to follow a monotonic evolutionary path. but
seems rather to go through a series of punctuated equilibria (from mobile phones to
portable phones and. more recently, web enabled phones.) This suggests that companies
need the ability to switch from one approach to the other, something few are capable of
doing.
The second study. which dealt with fashion apparel, and the uncertainty that comes from
changes in fashion trends, highlighted the role played by certain key individuals in the
development and dissemination of these trends within an industry. That role is best
described as one of interpretive boundary spanning (Piore, Lester, & Malek, 1997b).
Interpretive boundary spanning goes beyond the dissemination of information across
organizational boundaries. Well-connected suppliers played a role of interpreting and reinterpreting new ideas and developments from one client firm to the other, in a situation
where copycat solutions would not have been acceptable. In one case, the same supplier
who excelled at playing such an interpretive role also performed a more analytical
function of disseminating best-practice information among the client's various plants.
The third study focused on the biotechnology and medical devices industries. In this case,
uncertainty is related to scientific discoveries, technological development, and the FDA
approval process.'4 This case highlighted the open-ended partnership agreements one
biotech company entered into, in order to maintain its position at the center of a wide
F4DA: Food and Drug Administration,
part of the U.S. Health and Human Services Administration
(www.fda.gov).
35
R&D network, even though it did not know at the outset what scientific or technological
developments the partners would contribute. In the case of medical devices, one
interesting finding was the important interpretive role of "evangelist" played by thought
leaders in a particular medical field on the dissemination and acceptance of a new
technology (Piore, Lester, & Malek, 1997a).
Finally, some of the findings from this research were abstracted into a set of metaphors
intended to help managers deal with ambiguous and equivocal situations (Lester, Piore. &
Malek, 1998).
Theory Building and Case Studies
The research approach used in this dissertation is the one used successfully in the larger
project, namely, theory building using empirical case studies, as described by Eisenhardt
(Eisenhardt, 1989). That approach draws on established qualitative research methods
from grounded theory (Mintzberg. 1979; Strauss. 1987) and inductive case study research
(Yin, 1994), among others. It is particularly appropriate in this case given that the aim of
this research is not to test certain hypotheses. but to provide a fresh perspective on a topic
that has been previously researched and to develop a new understanding and new
categories grounded in the empirical data.
Theory building starts with the broad research questions posed in Chapter I and
summarized above, and proceeds through repeated iterations between theory formulation
and empirical data analysis. In this respect, the work reported in this dissertation fits
within the context of the larger study presented earlier. The theoretical framework and
constructs developed in the next chapter should be viewed as the next iteration on, or a
refinement and extension of, the framework initially developed for the larger study.
Similarly, the case studies reported in the subsequent chapters represent a subset of the
cases developed for the larger project.
The definition of"theory" as used in this dissertation (and the larger project) is an
expansive one, so as to encompass the wide range of theoretical representations from the
different paradigms used to study design and product development-'any
coherent
description or explanation of observed or experienced phenomena" (Gioia & Pitre, 1990).
36
Eisenhardt points out that theory building takes a positivist view of research in that "the
process is directed towards the development of testable hypotheses and theory which are
generalizable across settings." (Eisenhardt, 1989, p. 548). She contrasts theory building
with the ethnographically oriented research approaches proposed by Strauss (Strauss,
1987) and van Maanen (Van Maanen, 1988.) where the desired outcome is "a rich and
complex description of the specific cases under study", or what Geertz calls a "thick
description" (Geertz, 1973). The work presented in this dissertation combines aspects of
both. I present a set of hypotheses at the end of the dissertation. However, the theory that
emerges from this work is in the form of a classification scheme intended to help
researchers and practitioners gain a new perspective on their work, and an effective way
of achieving that goal is indeed by providing a rich and convincing description of those
empirical findings that instantiate the different types of practices. The case reports in this
dissertation are written with that objective in mind, that is, to provide a rich enough
description of certain aspects of a case or a certain anecdote to help "achieve
intersubjectivity about a set of observations and their interpretation within an experience
collective" (namely, researchers and practitioners) (Butler, 1997). 5
The dissertation then has both a theoretical component and an empirical component that
are intricately connected to one another, in the sense that they co-evolved. Although the
theoretical component, in the form of a taxonomy of design and product development
practices, is presented first, the reader should keep in mind that these categories emerged
as a result of the empirical findings from the case studies. In turn, the theoretical
framework illuminated the interpretation of the empirical cases. For that reason, I use
both the terms "typology" and "taxonomy" to describe the classification scheme
developed in the dissertation. Given the nature of theory building, it would be inaccurate
to distinguish between a conceptually derived and an empirically based approach in this
case (Bailey, 1994, quoted in Staudenmayer, 1997).
'5 "Theory building seems to require rich description, the richness that comes from anecdote" (Mintzberg,
1979).
37
Theoretical Framework
The theoretical framework presented in the next chapter was motivated by some of the
empirical findings from the cases in this dissertation, as well as the findings from earlier
case studies. From those cases, it became clear that the analytical and interpretive
classification was too coarse to adequately describe the range and subtleties of design and
product development practices.
The framework I propose draws on ideas from the fields of linguistics and the philosophy
of language. First, the division between the structural and functional areas of linguistics is
a particularly useful one (Akmajian, Demers. Farmer, & Hamish, 1995). Syntax and
semantics, two important sub-fields of structural linguistics, offer a good model or
metaphor for analytical practices that are closed and self-contained, which are well-suited
for organizing and exchanging information within well-defined, well-circumscribed
domains in which all the interlocutors share a common pre-understanding. Concepts from
pragmatics. the functional area of linguistics, and from hermeneutics are used to flesh out
the elements of interpretive practice described earlier by Piore et al. (Piore et al., 1994).
My sources for material on pragmatics are the texts on linguistics by Akmajian et al. and
by Mey [Akmajian, 1995 #199; Mey. 1993 #201; I also refer to the work of Giv6n for a
broader, more abstract understanding (Giv6n, 1989). In hermeneutics I rely on Palmer's
overview of the topic and on Gadamer's Philosophical Hermeneutics (Palmer, 1969;
Gadamer, 1976), with occasional references to Dreyfus' commentary on Heidegger's
Being and Time (Dreyfus, 1991). Pragmatics and hermeneutics provide rich descriptions
of what is involved in the process of coming to understanding in situations that require
the reliance on contextual factors, situations that are more open-ended and where the
interlocutors may come from different backgrounds or have different day-to-day
concerns. Table 111.3in the next chapter summarizes the elements that distinguish
analytical and interpretive practices.
Within the realm of the interpretive, I also distinguish between situations depending on
the role of the interlocutors and the purposiveness of the interaction, following Dascal
(Dascal, 1989).
38
EmpiricalResearch
The choice of industry
The case studies in this dissertation are drawn from a single industry. I chose to study the
automobile industry for a number of reasons. Chief among them is the high level of
complexity of the product and its interfaces, which are discussed at length later in this
section. Another reason is the level of maturity of the industry. This is an industry that
has seen a number of different trends in the last few decades, and one where a range of
organizational approaches and integrative mechanisms have been tried. A few that come
to mind: Ford's attempts to design a "world car" and to otherwise integrate its worldwide
design and development resources; GM's various reorganizations and its recent reliance
on "brand management"; Chrysler's use of dedicated platform teams and increased
reliance on its suppliers; the use of the tools and methodologies of the Total Quality
Management movement. This level of experience makes the industry a potentially rich
source of ideas and empirical evidence, an important requirement in theory building. A
related reason that makes the automobile industry interesting is the wealth of research
that has targeted it in the last several years, and the corresponding richness of secondary
research material available. A final reason is personal interest and a familiarity with the
product, having worked in the industry for several years.
The automobile industry is a good complement for the other industries studied in the
context of the larger project. Although it is subject to a host of regulations regarding
safety and pollution control, a new automobile design is obviously not subject to the
same level of regulatory uncertainty that is common in the pharmaceutical and medical
devices industry. Nor does it involve the uncertainty that accompanies the introduction of
a totally novel technology, as in the case of the cellular telephone industry. It does
however share one aspect with the apparel industry, namely, the uncertainty that comes
with shifting fashion trends. This goes beyond the obvious areas of interior and exterior
styling and color, to the area f product categories that seem to gain and lose popularity
for no predictable reason. Recent examples include the rise in the popularity of sport
utility vehicles or SUVs in the 1990s, and the sudden collapse of the market for sport
coupes earlier in that decade. (Bennet, 1995; Flint, 1996; Maskery, 1996).
39
An automobile, unlike an item of fashion apparel, is
a
very complex and expensive
product. Its design and development typically requires a relatively long lead time, ranging
from a year and a half to three years. Producing it requires expensive and specialized
tooling that must be amortized over a production run exceeding three years and tens of
thousands of copies. This makes it very costly to misjudge the market or make a design
mistake, and it puts a lot of pressure on the designers and developers to "get it right."
Despite this industry focus, the object of this dissertation is not to develop a theory of
product development practices specific to the automobile industry, but one that is more
broadly relevant. Towards a similar, end, Clark and Fujimoto present a framework for
describing product complexity along two dimensions, internal and external (Clark et al.,
1991, pp. 10-11). The first represents the complexity of the internal product structure,
which depends on the number of components that make up the product, on how many of
these interface with one another, the nature of the interactions at these interfaces and the
"severity of the tradeoffs" between the components (idem). The second dimension,
external complexity, reflects the complexity of the interface between the product and its
producer on one side, and the user of that product on the other. External complexity
varies with the number and specificity of the criteria the customer uses in judging the
product. It also depends on whether these criteria can be considered individually or
whether they are used in a holistic assessment, and on whether the criteria involve
dimensions that can be measured objectively or ones that are typically judged
idiosyncratically.
As an example of internal complexity. and the design challenges that go with it, consider
the problem of reducing the level of noise and vibration in a vehicle interior. The "NVH"
(for noise, vibration, and harshness) characteristics of a vehicle depend on the
characteristics of a large number of components, among them the following: the structure
of the vehicle, including its stiffness and damping properties (which may depend on such
an unlikely component as the anti-corrosion coating used); the engine, its architecture
(number and layout of cylinders), its internal mechanical components (such as the
camshaft drive mechanism), the design and tuning of its air intake and exhaust systems;
the fuel system including the fuel pump and its mounting hardware, and even the fuel
40
injectors; the arrangement used to attach the powertrain to the chassis, and the
characteristics of the mounts; the suspension design and its tuning, including the
characteristics of the springs, dampers, mounting bushings, and tires, most of which are
highly nonlinear; the door seals; the thickness of the windshield and other glass; the
properties of the materials used in the interior such as the carpeting, headliner, and
upholstery. These components involve complex and non-obvious interactions, with a
source of vibration in one component often setting off a resonance in another, such as
when engine vibration at idle causes the steering column to shake. Furthermore, many of
these components are involved in some of the more difficult tradeoffs in designing and
tuning a vehicle. The obvious one is the tradeoff between using additional amounts of
sound deadening material and the loss of performance that results from the increased
weight. Another tradeoff is the tuning of the suspension components: softer springs and
bushings generally reduce the level of NVH at the cost of less precise handling and
tracking. Similarly, softer powertrain mounts typically afford better isolation of engine
vibration, but they cause unpleasant jerkiness at the throttle tip-in and tip-off (when the
throttle is opened and closed.) Needless to say, cost plays an important factor in the
severity of these tradeoffs: more complicated and therefore more expensive systems such
as adaptive suspensions or adaptive intake and exhaust systems can always be used to
move from one tradeoff curve to another one that is more advantageous. In any case,
these situations are made all the more difficult by the fact that choosing a particular point
along any of these tradeoff curves is not a simple objective decision, but one that involves
subjective assessments by development engineers and managers, who rely on their
experience, knowledge of the competition's offerings, and their understanding of their
customers' expectations. Which brings us to the external complexity of the automobile.
Typically, few people choose between different automobiles purely on the basis of their
ability to provide basic transportation for a given number of people and a given amount
of luggage. An auto purchase is often an emotional decision, in which "fantasy and
imagery" play an important role (Clark and Fujimoto, 1991, op. cit.). Again, sports cars
and SUVs provide excellent examples. A typical modem two-door coupe offers few
advantages in terms of weight or aerodynamics, when compared to an equivalent fourdoor sedan. Yet, the typical buyer ascribes a higher level of performance to the two-door
41
version, as well as an image of freedom, playfulness and individualism. In terms of
performance, many high-performance sports car offer levels of acceleration and speed
that dramatically exceed what is usable on public roads, even by a competent and
experienced driver; yet, to many of their buyers. any less performance would seriously
detract from the value proposition these products present. Similarly in the case of SUVs:
research consistently indicates that most of their owners rarely take them off-road. SUV
buyers are not choosing these vehicles for their off-road capabilities, but, in many cases,
for the sense of security they perceive as a result of the rugged appearance of these
vehicles (Bennet, 1995; Malek, 1996). Interestingly, some the fantasy and imagery on
which some automobile buying decisions are based can be traced to a point in time when
the buyer was not even of driving age. Needless to say, much of this imagery is socially
constructed, with significant support provided by the marketing departments of the
manufacturers, through their various advertising campaigns. Obviously, such factors are
notoriously 'difficult to translate into technical specifications" for a new product (Clark
and Fujimoto, 1991. op. cit.). Clark and Fujimoto mention other aspects of the customers'
wants and needs that contribute to the complexity of the external interface of the
automobile. One is the fact that "car consumers are often unable to articulate their future
expectations, although they can tell which products they like when they see them." This
is particularly true when it comes to exterior and interior styling and color. In essence.
market needs in the case of the automobile are "subtle, latent, equivocal, and inarticulate"
(ibid.).
In view of the high level of internal and external complexity that is characteristic of the
automobile, it is reasonable to assume that designers and developers in that industry are
likely to have evolved approaches and practices that are interesting and worth studying.
The automobile industry seems to present fertile ground for theory building research.
Furthermore, to the extent that other products involve one or the other of the two kinds of
interface complexity described above, the types of approaches uncovered by this research
here for dealing with such complexity would be relevant.
42
The cases
Case selection in theory building does not follow the concepts of statistical sampling,
where the concern is finding several matched research sites where findings are likely to
be duplicated, thus lending them more validity. In theory building, research sites are
selected based on the notion of "theoretical sampling", in order to "provide examples of
polar types" for example (Eisenhardt, 1989). Selecting sites that are diverse decreases the
likelihood of developing theory or categories that are specific to one particular type of
organization. The three sites I selected are a US company in Detroit (Chrysler), the
Southern California design studio of a Japanese company (Nissan Design International,
or NDI), and the Japanese parent company itself (Nissan Technical Center, or NTC).
Aside from the difference in national origin and location, Chrysler and Nissan had
different fortunes at the time when the studies were conducted. After coming close to
bankruptcy, Chrysler was riding a wave of success (Taylor. 1994). Conversely, Nissan.
which had been relatively successful in the late 1980s, was doing poorly at the time
(Thornton & Armstrong, 1997).
Another criterion I used in selecting the study sites was that the companies and their
products be "interesting." I believe that each of the two companies selected qualifies.
Chrysler has been a design innovator in the industry, both with its mainstream offerings
and with the unusual specialty models introduced in the last few years. Examples of the
former include the original Neon subcompact, the Cirrus compact sedan, the LH sedans
with their cab-forward" styling (e.g.. the Dodge Intrepid). and the Dodge Ram Pickup
truck (Fleming, 1991; Bennet, 1994; Taras. 1994; Nussbaum. 1995; Flint, 1997).
Examples of the latter are the dodge Viper roadster and the more recent Plymouth
Prowler, two products that do not fit into any previously established product categories
(Connelly, 1990; Bohn, 1993; Keebler, 1993). Products aside, Chrysler is an interesting
study site from an organizational and management perspective, as it was one of the first
companies in Detroit to rely on dedicated platform development teams and to shift
significant engineering responsibility to its suppliers (Vasilash. 1991; Flint. 1992; 1997).
In addition, the management team at Chrysler, during the time when the study was
conducted, included a few interesting personalities, such as Bob Lutz, the chief operating
officer, and Robert Eaton, Chrysler's Chairman (Gelsi, 1996). These factors allowed the
43
company to achieve high levels of profitability without relying on radical cost cutting
measures, such as plant closings, like some of its competitors (Taylor, 1994).
Nissan has also produced some remarkable products, although none as radical as
Chrysler's Viper or Prowler. The mid to late 1980s was a prolific period for the company,
resulting in the original Nissan Pathfinder SUV. the 300ZX sports coup6. the Maxima
sporty sedan, as well as some of the Infinity models such as the original Q45 sedan and
the later J30 with its unusual ovoid shape (Armstrong. 1991; Bryant, 1992). Another
unusual design is the original Nissan Altima subcompact sedan, elements of which
appear inspired by the J30 (Clements, 1992; Taylor, 1993). Nissan was also one of the
first foreign automobile manufacturers to establish a design studio in Southern California,
a move that was followed by others. including the US manufacturers (Barron, 1998). NDI
is itself an interesting organization because it has always seemed somewhat independent
of its parent company, acting more like one of the well-known Italian design houses
(ItalDesign or Pininfarina) than the captive design arm of a large Japanese manufacturer.
It has undertaken a number of outside, non-automotive projects. and it is led by an
outspoken president who is well known in the industrial design community (Armstrong,
1991; Washington, 1995; WSJ, 1999). The designs developed at NDI are often in
competition with work done at the Nissan Technical Center in Japan (Armstrong, 1991.
op. cit.). where most of the engineering and development work is performed to
productionize these designs in any case. It is therefore only natural to complement the
NDI study with a study of NTC. NTC has also developed a few specialized models on its
own, some of which were not sold in the US (such as the be-I and Escargot minicars of
the late 1980s), and others than did not go beyond the prototype stage.
The design of the case studies follows Yin's embedded multiple-case format (Yin. 1994).
Each of the three cases involves several units of analysis. At one level, I use a particular
product to anchor each of the cases: a new Minivan project at Chrysler. and the second
generation Altima at Nissan. However, other products were discussed at some length
during the studies, such as the Viper and Prowler in the case of Chrysler. and a future
sports car and a small SUV in the case of Nissan. These were typically covered in the
course of studying particular units within these companies-the Design Office in the case
44
of Chrysler, and the Design Context Laboratory and Color Studio in the case of NDI. The
sub-cases include: design studios in Detroit, California, and Japan; a color studio in
California; a "Design Context Lab", also in California; a multidisciplinary platform
development team in Detroit; an interior supplier in Detroit. (For a complete list of the
cases and the interviews, please refer to Table 11.1 below.) I believe that the range and
richness of these embedded cases, which were inspired by the notion of theoretical
sampling, contributes to the relevance and range of applicability of the findings
Table /I. 1. The case studies sites
Site
Chrysler
Nissan Design Int'l.
Nissan Tech. Ctr.
Location
Detroit, Ml
La Jolla, CA
Atsugi, Japan
Minivan
Altima
Altima
(Dodge Viper)
(Future mini SUV)
(Various)
(Plymouth Prowler)
(Future sports car)
Product(s)
(LH sedans)
*
Design Director
(Corporate Design
*
President of NDI (I)
(2 interviews)
*
Design Studio
(Altima) heads (2)
·
Color Studio head ( )
·
Head of Exterior
Design Studio #2 (2)
*
Design Context
Laboratory manager
( I)
*
Head of Product
Experiment Dept.#2
(Marketability testing)
·
Office)( )16
*
Shukan & ex-Shukan
(now Department
Managerof Product
Minivan program
Design Administration
Dept.) (2)
manager (I)
*
Minivan team
managers and various
Executive Engineers
(4)
Sub-cases /
Interviews
·
Director of Advanced
(I)
Manufacturing
Engineering and Body
Engineering Executive
Engineer (2)
*
Supplier VP and
personnel (Lear
Seating) (5)
16
Number of respondents present at interview.
45
The main research instrument I rely on is qualitative, open-ended interviewing (Piore,
1979; Rubin & Rubin, 1995). Given the nature of the research questions, open-ended
interviews with design and PD managers and practitioners represent the most appropriate
approach. They allow the important issues facing the practitioners. and the ways in which
they conceive of them and approach them, to emerge during the interview. Given the
broad range of activities in design and PD, and the wide variation between the different
sites, different informants will have different areas of concern. making a more structured
approach inappropriate. This approach also makes it possible to capture the practitioners'
description of their practices using their own terms and distinctions. instead of forcing
their responses into predefined categories. as is likely to happen with other. more
structured research instruments (Mintzberg, 1979). The interviews were initially given a
general direction (market. customer. interaction with other functions or departments,
selection of design). but then were allowed to proceed according to the subject matter
uncovered. For the same reasons. I do not use any intensive detailed coding techniques
(such as computerized coding), which could risk destroying the meaning of the empirical
data (Eisenhardt. 1989, p. 534).
The reports and analysis of the findings at the three sites. Chrysler. Nissan NDI, and
Nissan NTC. are presented in Chapters IV through VI.
46
Chapter III:Theoretical Framework
A Taxonomy of PD Practices Inspired by Linguistics and the
Philosophy of Language
This research is concerned with the interpretive communicative processes and practices
which people involved in design and product development deploy when interacting with
their customers and with each other. The aim is to gain a better understanding of these
processes and to develop a vocabulary that would make them more understandable and
more accessible to the PD practitioner.
Interpretation is an activity that humans engage in constantly as they face the countless
situations that make up their daily lives. In the introduction of his book on hermeneutics,
Palmer describes "the ubiquity of interpretation" and asserts that living, at least for
humans, is itself a "constant process of interpretation"(Palmer,
1969, pp. 8-9).
Considerfor a moment the ubiquity,of interpretation.and the generalityof
the usage of the word. The scientistcalls his analysisof data
"interpretation"; the literarycritic calls his examinationqf a work
"interpretation." The translatorof a languageis called an "interpreter";
a news commentator "interprets" the news. You interpret-or
misinterpret-the remark of afriend, a letterfrom home, or a sign on the
street. In fact, from the time you wake in the morning to the time you sink
into sleep,you are "interpreting." On waking,you glance at the bedside
clock and interpret its meaning:you recall what day it is, and in grasping
the meaningof the dayyou are alreadyprimordiallyrecallingto yourself
the wayyou areplaced in the world andyour plans for thefuture; you rise
and must interpretthe wordsand gestures of thoseyou meet on the daily
round. Interpretationis, then,perhaps the most basic act of human
thinking;indeed,existingitself may be said to be a constantprocess of
interpretation.
47
Consider in particular the case of a designer interacting with potential customers for a
product he is working on, for the purpose of learning and assessing what these customers
want from the product, what they expect, what they might or might not accept, what they
are likely to find pleasantly surprising or totally unresponsive to their needs. Whether the
interaction is direct or indirect-in other words, whether it involves reading a marketing
brief, studying the results of a survey, or talking directly to a group of customers-the
designer is constantly interpreting what he is reading or hearing. Similarly, as designers
and product developers interact with one another in order to understand what one
functional group needs or expects from another, what they might be willing to accept or
able to change, and how they might be convinced to do so, they are also constantly
interpreting what they are being told by their interlocutors. If we accept, provisionally,
that interpretation may be the "most basic act of human thinking", them it must follow
that interpretation is pervasive in a creative activity such as product development.
Although the interactions mentioned in the previous paragraph are described in linguistic
terms. interpretation itself is not limited to the linguistic world. For example. Palmer
describes whatever process an animal undergoes as it sees a piece of food in front of it as
interpretation; and he describes migratory birds interpretation of a variety of inputs as a
sign that they should fly south (Palmer. ibid.). Returning to humans, we all know how to
interpret a hand gesture as a greeting. or a certain look as a sign of anger. In reality, as
designers interact with customers and with one another. they rely on a number of
different media. Most often, the discussions and conversations in question are not limited
to linguistic exchanges. They also involve such things as pictures, sketches, drawings,
graphs, data tables and plots, as well as models and prototypes. whether physical or in the
form of computer generated representations. For example. an automotive consuner clinic
will typically feature several current production vehicles as well as one or two prototypes
of the proposed model. And it would be hard to imagine engineers from different
functional groups discussing a design without looking at some drawings, whether over a
drawing board or a CAD screen. Some of these artifacts play a crucial role in bringing the
interlocutors to a common understanding. and they have been the subject of various
research projects, such as the study by Carlile on the role of boundary objects in product
development (Carlile, 1997).
48
The theoretical framework developed in this chapter is derived from linguistic concepts.
Still, the claim here is that the presence of these non-linguistic elements in the
interactions in question, and the importance of the role they play, do not detract from the
relevance and applicability of the framework. One way to make this claim reasonable is
to point out that although the framework does indeed start with linguistic concepts. it
ultimately deals with interpretation in general, which has universal relevance as was
pointed out earlier. But a better explanation would be to point out the importance and
universality of language itself to our existence as human beings, and the peculiar role it
plays in our knowing and understanding.
Palmer argues that language shapes how man sees his world, how he thinks, and even
how he feels; then, using ideas from Heidegger and Gadamer, he describes language as
"the 'medium' in which we live, move, and have our being" (Palmer. ibid., p. 9). He also
discusses Heidegger's increasing emphasis, in his later writings, on "the linguisticality
(Sprachlichkeit) of man's way of being" (that is, on man's way of being human, or
"Dasein") (Palmer, ibid. p. 155).
Gadamer discusses the inseparability of human knowledge and language in his essay on
"Man and Language":
... in all our knowledge of ourselves and in all knowledge of the world we
are always already encompassed by the language that is our own. We
grow up, and we become acquainted with men and in the last analysis with
ourselves when we learn to speak. (Gadamer, 1976, p. 62).
And later:
Hence language is the real medium of human being, f we only see it in the
realm that it alone fills out. the realm of human being-togetherness, the
realm of commonunderstanding,of ever-replenishedcommon
agreement... (Gadamer, ibid., p. 68).
On a less philosophical level, researchers in cognitive science and linguistics have written
extensively on the connection between language and thought. For example, such a
seemingly purely linguistic construct as metaphor turns out to be pervasive in the way our
thoughts and actions are structured (Lakoff et al., 1980). Similarly for metonymy, another
49
construct, which is not only linguistic but cognitive as well, since it is found to underpin
an important area of how humans combine things in categories (Lakoff, 1987).'7
In the next section, I introduce the basic areas of linguistics, focusing in particular on
syntax, semantics, and pragmatics. (Although I have used hermeneutic ideas from
Gadamer and Heidegger to motivate the focus on language, there is no inconsistency in
turning now, temporarily, to some of the more analytical, structural aspects of language
study.) I then introduce, in another section, some concepts from hermeneutics,
occasionally comparing pragmatic interpretation to hermeneutic interpretation. In the
third section of this chapter, I present and explain the classification framework proper, as
well as an example of how it may be used.
Background I: Linguistics
Linguistics. the study of human natural language and communication, is a far ranging
field. Its concerns include such disparate areas as the origin of words, the rules used to
construct proper sentences, the physics of sound formation through the vocal chords, the
study of brain function during speech, and the evolution of dialects in different parts of
the world. Here, we are interested in language proper. not in its physical, neurological. or
ethnological aspects. In general, the study of language is divided into two broad
categories, the structural and the functional.
Structurallinguistics8
Structural linguistics comprises the sub-fields of morphology, phonetics, phonology.
syntax, and semantics. Akmajian et al. note the common theme that runs through these
sub fields:
... the structuralanalysisof human languagecan be stated in termsof(l)
discreteunits of varioussorts and (2) rules and principles that govern the
way these discrete units can he combined and ordered (Akmajian et al..
1995. p. 3).
'7 Metonymy refers to the situation where the name of an object or concept is used to refer to another to
which it is related or of which it is a part (Random House. 1992).
's
The presentation of structural linguistics draws heavily on the textbook by Akmajian et al.. (1995).
50
These sub fields are concerned with developing descriptive rules, that is, regularities and
generalizations observed in the structure of language, not with prescriptive rules on the
proper usage of language, such as the rules of grammar.
An important aspect of structural linguistics is that although a finite number of rules and
principles are used to describe the system, it nevertheless remains unbounded in scope. In
other words, there is no limit to the new words or sentences that can be formed, or the
kinds of things that can be talked about. This principle of "effability" points out a
similarity between language and design, namely, the unboundedness of the range of
products of what is basically a creative human activity.
Morphology is the study of words and the basic building blocks they are made of; how
these elements are put together to form more complex words, and how the meanings of
these more complex words are related to the meanings of the parts, and the relations
between words.
The field of phonetics deals with how speech sounds are produced. focusing on their
production in the vocal tract in the case of articulatory phonetics, or on the physical
properties of the sound waves thus produced in the case of acoustic phonetics.
Phonology deals with the patterning of sounds in natural languages, either comparing the
structures and patterns of sounds between languages, or looking for universal properties
that are common to natural language sound systems.
Of more interest to us are the two remaining areas of structural linguistics, syntax and
semantics, because they focus on phrases and sentences, the basic buildirngblocks of
communication, instead of looking at sounds or words only.
Syntax
Syntax is the study of phrases and sentences, the basic elements of linguistic
communication. Here, the focus is on the patterns of formation of phrases and sentences
from words, and on the rules used in the formation of grammatical sentences in a
language. Since the publication of Chomsky's seminal work on the subject, "Syntactic
Structures", in 1957, the field has been dominated by the transformational grammar
paradigm (Chomsky, 1957).
51
A grammar consists of a finite set of rules and principles that reflect the regularities
found in the language, and that enable a speaker "to produce and comprehend the
unlimited number of phrases and sentences of the language" (Akmajian et al., op. cit., p.
137). It should be noted that "comprehend" is used here in the sense of making
grammaticality judgments about sentences, not in the sense of understanding the meaning
of these sentences; that falls in the domain of semantics, which is covered later.
A critical element of any grammar is a lexicon that provides a classification for the
building blocks of the language, i.e.. the words of the language, into specific categories.
To get an idea about the kinds of rules involved in a theory of syntax, we can look at an
informal theory of syntax proposed by Akmajian et al. (op. cit.). In that introductory
theory, the authors identify a few basic aspects that are reproduced and explained below.
The first three deal directly with the internal structure of sentences:
Structuralproperties:
a. The linear ordering of elements. This property reflects the fact hat the ordering of
words or elements (e.g., noun phrases) in a sentence is not arbitrary. Changing that
order will affect whether the sentence is seen as well-formed or not, grammatical or
not; to say nothing of how a change in the ordering might affect what the sentence
may mean;
b. The labeling of elements into morphological or lexical categories (also known as
"parts of speech".) This labeling resolves whether a particular element in a sentence is
a noun, a verb. an article, and so on. These lexical entries that are needed for every
word in the language are separate from the meaning or denotation of the word. (These
aspects will be covered in the next section. which deals with semantics.)
c. The grouping of elements into structural constituents or phrases. A sentence element
can itself be a linear grouping of other elements; for example, a prepositional phrase
can consist of a preposition and a noun phrase which itself is composed of an article
and a noun.
Another set of elements of the grammar presented in this informal theory relates to how
certain phrases function. grammatically, within a sentence. Examples of these are the
52
concepts of subject and object in a sentence. At first sight, these relationships may seem
to fall outside the purview of mere syntax, as they require the speaker or hearer to make a
judgement concerning who is undertaking the act described by the verb and who is being
subjected to it. However, these grammatical relationships can indeed be reduced to
structural descriptions (with some help from the lexical entries for the verbs) as follows:
Grammaticalrelations:
a.
Subject (a noun phrase that precedes a verb phrase, within a given sentence.)
b. Object (a noun phrase that follows the verb in a verb phrase.)
What we have then is a self-contained system that is purely structural in nature. It allows
us to decide, given a sentence, whether it is correct or not. It also allows us to generate
any grammatical sentence in the language using a finite set of transformation rules. It
does not tell us anything, however, about what these sentences might mean. And the
process of generating these sentences can be purely mechanical; it does not require any of
the mental abilities of a thinking human being, or any intentionality. 9
Before moving on to semantics, it may be useful to look at what a formal theory of syntax
looks like. A basic phrase structure grammar (generative grammar) would consist of a
number of phrase structure rules (PS rules), such as the following:
S
NP Aux VP
NP - (Art) N
VP - V (NP)
Where S refers to sentence, and P to phrase. N, V, Aux, and Art refer to the lexical
categories noun, verb, auxiliary, and article, respectively. And the parentheses indicate an
optional constituent. The arrow is used to indicate a "consists of' relationship or rule. The
first rule indicates that a sentence S consists of a noun phrase NP, followed by an
auxiliary Aux and by a verb phrase VP. And the second relationship gives the rule for the
formation of a noun phrase NP: it consists of an optional article (Art) and a noun.
'9 Intentionality is used to refer to "things that are about other things" such as mental states and beliefs, but
also, indirectly, sentences(Audi, 1995).
53
The importance of the lexical categories is clear from the relation between lexical
categories and phrasal categories, namely: For X, X a lexical category (N, V, A, P), we
can form a phrase of type X with X as its head, such as: XP -) ... X ...
Semantics
Semantics is concerned with the study of meaning and reference in language. It is
intended to provide an account of what linguistic units-phrases and sentences-mean,
and what they refer to. There is a difference in emphasis between semantics as used in
linguistics, and as it used in logic. Reference, that is, denotation and truth conditions,
finds greater emphasis in the latter.
Until the early 1960s, the inclusion of semantics in a grammar, that is, a system of rules
specifying a language, was considered "not quite respectable" (Akmajian et al., op. cit., p.
213), and the "waste-basket of linguistics" (Mey, 1993, p. 5). That is no longer the case:
linguistic competence is now viewed as including meaning, and semantics, like syntax, is
now considered a legitimate part of a grammar.
When dealing with the issue of meaning, a basic distinction must be made between
linguistic meaning and speaker meaning. The former is the literal meaning of an
expression, that is, its strict, non-figurative meaning in the language. The utterer's or
speaker's meaning, on the other hand, may or may not match the linguistic meaning,
depending on whether the speaker is speaking literally or not. Examples of nonliteral
meaning are those involving irony or sarcasm, as well as the use of metaphor.
Beyond this basic distinction, the field of semantics has had to deal with the lack of a
consensus among researchers as to what meaning really is. Several theories of meaning
have been proposed over the years. The easiest ones to grasp are those theories that
conceived of meaning as an entity. For example, in the Denotational Theory, the meaning
of an expression is the actual object denoted by that expression. In the Mentalist Theories
of meaning, the relevant entities are the ideas that are associated, in the mind of the
speaker, with the expression in question. In the Image Theory of meaning, it is mental
images. All these theories suffer from various breakdowns under different conditions.
54
To deal with the problems stemming from the conception of meaning as an entity, Frege
introduced the idea of meaning as sense. The meaning of a sentence became its
satisfaction condition, that is, the condition that makes the sentence true (Akmajian et al.,
op. cit., pp. 221-222).
Later on, Wittgenstein introduced the Use Theory of meaning, which held that the
meaning of an expression is determined by how it is used in the community that uses the
language in question, the meaning of an expression being specified by specifying its use
(idem., p. 223). This conception of meaning is more useful and less limiting than the
others, but it comes closer to notions that are the purview of Pragmatics, which are
discussed later under functional linguistics.
These are only a few of the conceptions of meaning that have been proposed over the
years, and that have served as the basis for various researchers and their semantic
theories. For more detailed treatment, the reader is referred to (Audi, 1995, pp. 471-476).
Although Akmajian et al. (op. cit., p. 223) state "... it is fair to say that researchers do not
have a very clear idea what meaning is." it is still possible to lay out a set of issues of
concern in semantics, and to describe the elements that a semantic theory must explain or
account for.
One of the most important elements in a semantic theory is a lexicon. A semantic lexicon
goes beyond the one described under syntactic theory. Whereas the latter need only
provide "part of speech" information for every word (i.e., is a noun, is a verb, and so on),
a semantic lexicon must somehow represent the meaning of all the meaningful words of
the language. For example, such a lexicon would tell us that brother means male sibling.
It would also identify words that are synonymous, words that have several related
meanings (polysemous). With that information, it becomes possible to identify
expressions that are lexically ambiguous, redundant, or anomalous, as a result of
containing a word that is ambiguous (e.g., she can 't bear children), a word that already
includes the meaning of another (e.g., myfemale sister), or words that are incompatible
(e.g., he slowly ran up the stairs), respectively.
A critical element in understanding the meaning of a sentence is determining whether it is
an assertion, an order, or a question. A semantic theory should therefore be able to
55
account for the communicative potential of sentences: is the question a declarative, an
imperative or an interrogative sentence? Sentences of these different types have different
structures, and that structural information is therefore part of a semantic theory.
A semantic theory must also account for certain truth properties and truth relations of
sentences, such as entailment, where the truth of one sentence guarantees the truth of the
sentence that it entails. An example, from Akmajian et al. (op. cit.): the sentence The car
is red entails the sentence The car has a color. Another relation is semantic
presupposition, where if a presupposed sentence is false, then the presupposing sentence
cannot be said to be either true or false. For example, saying that My watch isfive
minutes too slow presupposes that I have a watch or that My watch is running; if either of
the latter statements is false, then one can no longer say whether my watch is or is not
slow.
At this point, the reader might be wondering about the connection between such detailed
and specific examples of linguistic semantics, and communication and knowledge in
product development. These examples are needed to show the reader what can seemingly
be accomplished with a grammar. a closed system consisting of a finite number of rules
and a lexicon. A semantic theory can tell us that a red car has color without ever having
experienced or seen anything that is red, purely based on its rules and the entries in its
lexicon. In other words, one can program a machine with the rules and lexical entries of
semantics, and that machine could then figure out that a particular sentence is a question
or a constatation, whether two sentences are synonymous or not, and that Chomsky's
famous "colorless green idea..." is a "meaningless" expression. "Meaning" in semantic
theory is therefore a highly specialized and limited version of what the average thinking
human being might understand by that term.
Functionallinguistics
In structural linguistics, human language is viewed as an abstract coding system having a
number of structural properties, such as the two levels briefly described above. syntax
and semantics. Functional linguistics is concerned with how languages are used by
humans, that is, how people use language to communicate in real life situations.
56
Specifically, we will look at Pragmatics, "the study of language use in relation to
language structure and context of use" (Akmajian et al., p. 340).
Pragmatics
As we saw earlier, the focus in semantics is on the linguistic meaning of phrases and
expressions, that is, their literal meaning. Very often in normal language use, however,
literal meaning and speaker meaning do not coincide, because people speak nonliterally.
such as when they communicate using sarcasm or metaphor. In pragmatics the focus is on
speaker meaning. The interest in pragmatics increased because of a sense of
dissatisfaction with structural linguistics as being too removed from real world language
use (Mey talks about "the collapse of earlier theories and hypotheses. in particular of the
'pan-syntacticism'
of Chomsky and his followers (Mey, op. cit., p. 19).) For example,
syntax and semantics tell us that ambiguity is rampant, due to structural ambiguity (the
structure of a sentence allowing for different interpretations, as in The mother of the boy
and the girl will arrive soon), as well as lexical ambiguity (a word having different
meanings, as in I can 't bear children.) Yet, in real life language use, ambiguity is rather
unusual, if we do not count those cases that involve purposeful evasiveness. In both of
the above examples, it is more than likely that the hearer will know exactly which of the
possible meanings is operative, from the context in which the expressions are uttered. It is
practically impossible to imagine a real situation in which these expressions would show
up-be uttered-in complete isolation, the way they appear above.
Another, more technical example of the shortcomings of semantics is the case of
presuppositions, briefly mentioned in the previous section. Mey gives the following
example. loosely paraphrased here (Mey, op. cit., p. 29). Consider the following two
sentences: John regrets that he failed the exam and John doesn 't regret that he failed the
exam. Regret being a factive verb, it presupposes the clause that follows it; in other
words, whether he regrets it or not, the presupposition in both sentences is that John has
failed the exam. As a result, on a semantics account, the sentence John doesn 't regret that
he failed the exam, because in fact he passed! would involve a logical contradiction. Yet
one can imagine a situation in which this last sentence might be uttered without sounding
illogical. Consider the scenario where one person might have just stated that John did
57
poorly on the exam, to which another responds by saying that John must regret failing the
exam; the first person could then utter the semantically questionable sentence, and it
would make perfect sense. As Mey points out, in this situation, the false presupposition
was the second person's alone, not the first person's (the speaker's) too, and it is
therefore not a semantic presupposition but a pragmatic one. It is important to note that,
here again, the scenarios in which that expression would make sense involve interactions
between speakers, not the isolated utterance of a sentence.
Akmajian et al. tie the limitations of the pre-Pragmatics view of communication to the
Message Model of communication, effectively the same as the model of communication
as information transfer that was discussed in the introduction (Chapter I) (Akmajian et
al., pp. 346-348). According to that model, the speaker typically has some message in
mind that he wants to communicate to the hearer: the speaker therefore encodes that
message into language and utters the corresponding sounds. The hearer receives the
message by hearing these sounds and decodes the utterance. identifying its syntax and
meaning using his knowledge of the language. The hearer finally recomposes the
meanings received into the original message. According to that model, successful
communication takes place when the message decoded by the hearer is the same message
that was initially encoded by the speaker. In the Message Model of communication,
language is seen as a bridge for ideas or. to use the common metaphor of the
communication channel, language is a "conduit for ideas" (Reddy, 1979).
The authors then identify a number of limitations of the Message Model as a
representation of human communication. First. there is the previously mentioned problem
of disambiguation, for which semantics does not account, as it offers no rules or
principles that might govern it. And yet, the fact that most people seem to manage the
process successfully most of the time indicates that it cannot be random, that it must
follow some principles. 2? One such principle is the presumption by the hearer of the
:" The shift from using the term "rules" in describing the regularities of the language. to using the term
"principles"., is connected to the shift from a focus on generative transformational grammars. In a
grammar, rules are supposed to "predict which sentences are correct", and to "establish ('generate') the
entire set of correct ('well-formed') sentences of a language..." Starting with semantics, and more so in
pragmatics, rules became "rules of usage, not of prediction." The focus on meaning, and in particular on
speaker meaning. make the concept of well-formedness
58
"controversial" (Mey, op. cit., pp. 54-55).
contextual appropriateness of the utterance. A second limitation stems from the fact that
meaning often is not sufficient to determine what an expression is referring to, and yet,
successful communication requires the hearer to determine what that reference is. The
example given in Akmajian et al. is that of the phrase "shrewd politician", which always
means the same thing, but which could be used on one occasion to refer to, say. Winston
Churchill, and on another to Richard Nixon. A third limitation is again one where
successfully decoding the meaning of the utterance is not sufficient for successful
communication; in this case, it is the communicative intent that is underdetermined. For
example, the utterance "I'm coming tonight" could be a threat, a promise, or simply a
prediction. In all three cases, the meaning of the utterance is the same, but the speaker
obviously intends to communicate different things. A fourth limitation of the Message
Model is that it does not account for nonliteral communication, where the speaker means
something other than the literal meaning of his utterance, as in the case where sarcasm is
used. Another case where communication involves more than decoding the meaning of
an utterance is indirect communication. For example, saying "l'm hungry" when what I
really want to communicate is my desire to eat. Finally, the sixth limitation of the
Message Model is that it does not account for those cases where the purpose of an
utterance is not to communicate a message, as in institutional speech acts or
perlocutionary acts. For example, when a boss tells an employee "You're fired", the
institutional act of changing the employment status of the employee is arguably more
important than the act of communicating it to him. And in the case of perlocutionary acts.
the speaker may say certain things to persuade the audience of something or to deceive
that audience; unlike communicative acts, where the communicative intention is intended
to be recognized by the hearer, in this case the speaker hope to hide his intentions.
The various thrusts of research in pragmatics came in response to these limitations. The
common thread in all these areas of pragmatics is the role of the context in
communication. One example is the inferential theory of communication of Bach and
Harnisch (Akmajian et al., op. cit.. pp. 352-368). That theory holds that 'linguistic
communication works because the speaker and hearer share a system of inferential
strategies leading from the utterances of an expression to the hearer's recognition of the
speaker's communicative intent" (Akmajian et al., op. cit., p. 352). The interlocutors
59
however share more than strategies; they also share a "system of shared beliefs." Instead
of the rules and conventions of a language or code connecting sounds and messages,
pragmatics proposes "systems of intended infeirenceand shared belief" or presumptions
that are at work during real life communication, and that we learn "in the co'urse of
learning to speak our language". A basic presumption is the Linguistic Presumption,
whereby 'the hearer is presumed capable of determining the meaning and referents of the
expression in the context of utterance." Another is the Communicative Presumption;
here. "the speaker is assumed to be speaking with some identifiable communicative
intent" unless there is evidence to the contrary. A third is the presumption of literalness,
by which, "unless there is evidence to the contrary, the speaker is assumed to be speaking
literally." These shared presumptions result in the hearer and speaker having a "fairly
specific set of conversational expectations" that enable the successful operation of the
shared system of inference strategies. These strategies are used by the hearer to determine
whether the speaker is communicating directly, literall) or nonliterally, or indirectly. In
each case, the test that helps the hearer choose the appropriate strategy is a test of
contextual appropriateness. To use an example from Akmajian et al. (op. cit.. p. 364). it
would be strange to assume that someone who just pulled into a gas station and reported
that he wants ten gallons of gas is not actually. in so doing, making an indirect request for
those ten gallons of gas.
Another area of pragmatics deals with speech acts, or acts that are performed by uttering
expressions, such as the perlocutionary acts mentioned above. In addition to
perlocutionary acts, where an act is performed by uttering something (as when one is
attempting to intimidate or persuade). speech acts include illocutionary acts. in which an
act is performed in uttering something (such as when one is promising or asking
something), among others (Akmajian et al.. op. cit.. pp. 376-380). In this area of
pragmatics. the aspect of context that is the focus of attention is the purposive context.
that is, the goal or purpose of verbal acts (Giv6n. 1989. p. 27). The complexity of the
issues involved in speech act theory has resulted in this area making significant
contributions to pragmatics, such as Searle's theory of the Background (with a capital B
to indicate the use of the term in a specific technical sense.) The Background reflects the
complex framework of social and institutional conventions within which we grow up as
60
humans, and upon which we draw when interpreting performatives and other speech acts
(Searle, 1985).
Pragmatics, more generally
Pragmatics as a perspective
In the previous section, pragmatics was generally presented as a branch or sub-area of
linguistics, just like morphology, syntax, and semantics. Many linguists do indeed view
pragmatics that way. Others, however, see pragmatics as a wholly different perspective
on linguistics. For example, Mey talks of a pragmatic perspective on phonology "that will
emphasize the societal aspects that are inherent in a certain phonological system" (Mey,
p. 46). This distinction goes beyond a matter of disagreement about the boundaries of one
or another sub-area of linguistics, and it is not of one sub-area involving more than one
level of analysis. (In semantics for example, one can look at the meaning of a morpheme,
a word, a phrase, or a sentence, as the principle of compositionality connects the latters to
the formers.) Mey traces the view of pragmatics as another of the branches of linguistics
to the "component view" of linguistics, popularized by Chomsky and the researchers who
followed him in the field of generative transformational grammar. The "component view"
holds that the mind has a number of distinct, independent but cooperating, faculties (Mey,
p. 45). Along that view, each of the components or "modules" of linguistics "works
within a properly delimited domain, with proper, well-defined objects, and with properly
established, specific methods" (Mey, p. 46). From the description of the previous section.
one can see that it may be difficult to clearly circumscribe the domain of pragmatics, to
delimit the notion of context, or to think of it in terms of "well-defined objects." (Mey
quotes Ostman, a pragmatician, who describes the unit of analysis for pragmatics as
being the whole 'functioning of language".) Alternatively, the two views could coexist
side by side, as "they both are metaphors, designed to expand, not to narrow our
epistemological horizons" (Mey, p. 47).
The universality of pragmatics;framing it using systems theory and logic
Pragmatics is more than a component of linguistics or even a perspective or "umbrella"
for the study of linguistics and language use. Giv6n defines pragmatics as follows:
61
Pragmaticsis an approachto description,to informationprocessing, thus
to the construction, interpretation and communication of experience. At its
core lies the notion of context, and the axiom that reality and/or
experience are not absolute fixed entities, but rather frame-dependent
contingent upon the obserer
er perspective (Giv6n, 1989,p. xvii). 2 1
Giv6n traces the roots of pragmatics from "pre-Socratic Greek dialecticians, then via
Aristotle to Locke, Kant and Pierce, eventually to
1 9 th
Century phenomenologists"
and
on to Wittgenstein (ibid.).
In Giv6n, the role of pragmatics in relating the meaning of an expression to its context is
presented as only one of different ways of looking at the role of perspective or frame of
reference, and it is presented as a metaphor. Other metaphors he uses: description and
point of view (a description being incomplete ("uninterpretable") if it does not include
the point of view from where it is taken); picture and frame (a picture being not fully
specified unless its frame is also specified.") The author explains the role of pragmatics
using notions from systems theory and logic, and he frames the issue in terms of two
predicaments. His explanation is paraphrased here.
The three core metaphorsfor the pragmaticmethodgiven above-point of
view,frame, and context-may be.firther generalizedvia the notions of'
systems and meta-levels.[...] Thesystem,as a hierarchicentity, is made
out of a progression oflevels. each one acting as a meta-level[...] thus
the contextfor the sub-levels embedded within it. For purely practical
reasons, f the system is to remainfinite (i.e. describable withinfinite time,
space and means), the last-highest-meta-level must remain contextless; it lacks its own meta-level. In terms of our picture metaphor, the last
meta-level is the.frame, yet itself remains un-framed, therefore, not fullyspecified (Givon, op. cit., pp. 2-3).
21
The use of the expression "information processing" may be confusing here, as information processing is
generally thought of as an analytical, structured process that is more akin to the descriptions of
transformational-generative grammars. Often, authors will set up the analytical information processing
view in opposition to the more interpretive. practice oriented views of human thought and behavior (see,
for example, Carlile, 1997.) There is no reason why, however, interpretation cannot be viewed as a kind
of information processing. as in the above quote. In that case, information processing is a more general
use of the term, whereas the information processing view of communication or interaction should be
understood as a particular perspective on these types of activities (Carlile,, 1999). That view fits within
the reductive, logico-deductive view discussed later.
62
This leads to what Givon calls the first predicament of pragmatics, the predicament of
completeness, which states that "So long as a system is fully specified, i.e., closed, it
must remain in principle incomplete."
This framing of pragmatics using systems theory is useful; for example, it can easily, and
in a clear fashion, account for a previously mentioned characteristic of semantics that
seems paradoxical, namely, the fact that every word in the lexicon is explained using
other words in the lexicon. Such a "closed", "fully-specified" system might have
intuitively struck the reader as lacking. as being circular or somehow 'incomplete".
Another way of looking at the problem of completeness from a systems theory
perspective is to focus on the inputs and outputs at the boundary of a system. When
analyzing a system, one would normally pick a level of analysis and decide on the parts
of the system. or the aspects of its behavior, that are to be included or modeled, and those
that are to be left out. The result of this selection process is a definition of the system, its
boundary, and its environment (technically, everything in the universe that is not part of
the system.) The problem then is that all interactions between the system and its
environment must be fully characterizable as a finite, well-defined set of inputs and
outputs across the boundary. In the case of physical systems, energy considerations (or
frequency separation considerations) offer crucial guidance to the analyst in deciding
what to include inside the boundary and what to leave out, and in deciding which inputs
are significant and which can be neglected. No such considerations exist in the case of
information systems, and especially in those cases where information is to be processed
in non-mechanistic ways. For example, in trying to explain a concept, or when trying to
create a novel design. a small, seemingly insignificant. piece of information could offer a
radically different way of framing the situation or problem, and therefore a totally new
explanation or design. (The idea of assigning some measure of importance to
information, based on the contribution of that information in reducing uncertainty, as
briefly discussed in Chapter 1,can be understood as an attempt to develop for information
something akin to energy.)
The second element in Giv6n's motivation of pragmatics is the predicament of
consistence which he states as: "So long as one is allowed to switch meta-levels--or
63
points of view-in the middle of a description, the description is logically inconsistent."
Giv6n traces this to Bertrand Russell's observation "that the classical paradoxes of logic,
such as the liar ' paradox ('I never tell the truth') are all instances of a more general
phenomenon. that of self-inclusion." By constraining logical descriptions to a single
specified meta-level, Russell eliminated deductive logic "as a serious contender for
modeling, describing or explaining human language-or mind", in the process of
"rescuing" it "as a closed, internally-consistent, coherent system" (ibid.).
But the problem is not deductive logic per se; the predicaments of closure and
consistence present serious difficulties for all attempts at describing language and the
human mind. Giv6n explains:
Neither language nor mind abides by the requirements of'closure, except
perhaps temporarilyjbr limited tasks. Both mind and languageare
necessarilyopen systems that constantlyexpand add meta-level,learn
and modify themselves. Equally, both language and mind are notoriously
promiscuousin violatingRussell's constrainton self-inclusionand
reflexivity.('onsciousnessis indeedlbrever adjustingils.frame,shifting
meta-levels;:it keeps re-framing and reflexively framing itself This
propensity [...] is a preconditionjibr the mind 's ability to select, evaluate.
file, contextualizeand respondappropriatelyto mountainsof inbfrmation.
More on context
It should be clear, from the previous discussion, that context in pragmatics is not an
easily circumscribed or fixed entity. Akmajian et al. describe the context of an utterance
as "an expandable notion" (Akmajian et al.. p. 370). Depending on the particular situation
and on the utterance itself, the relevant context could simply consist of the preceding
utterances, along with the few ones that are yet to come. This is referred to as the co-text
of the utterance(s). Mey defines "the co-text of a (single or multiple) sentence as that
portion of text which (more or less immediately) surrounds it." And he points out that
there is "no agreed limit" as to how far it extends (Mey, p. 184).
Beyond co-text. context can extend to the "immediate physical and social environment"
within which the sentence is uttered (Akmajian et al., ibid.). In other words, it could
include details about where and when the utterance took place. who was present there.
what the weather was like, and so on. Such contextual information would be useful in
64
resolving ambiguities related to the use of deictics (such as he, we, here, etc.) for
example. It would help resolve questions as to whether an expression such as Nice
weather! was meant sarcastically or not.
The third level of context encompasses all kinds of general knowledge, such as that
relating to general social and institutional practices. The following two examples will
help clarify what kind of knowledge is embodied in this level of context. The first is an
example from Levinson (reproduced in Mey, p. 186). and it involves the following
exchange:
A. I have a fourteenyear old son
B: Well that 's all right
A: I also have a dog
B: Oh I'm sorry
This exchange would not be properly understood unless one knew about such social and
institutional facts and practices as property, housing rental, rules that some property
owners may want their tenants to abide by, and so on.
The second example paraphrases one given by Searle as he lays out his theory of the
Background. Consider the following exchange:
A: One cheeseburger and a small firies
B: That will befour ninety-nine
In order to understand this exchange, one has to know, among other things. about
restaurants, fast food restaurants in particular, about what constitutes food, about the
practices of buying and selling, about the institution of using paper money. what value is.
and so on and so forth.
Depending on the situation, the outer level of context may involve knowledge that is not
only general but also, paradoxically, somewhat particular to the practice within which
that situation happens to fall. For example, the first example above is tied to the practice
or real estate and housing rental; the second to the practice of eating out. It is easy to
65
imagine exchanges that fall within more specialized practices, and that would therefore
require the specialized knowledge that is part of these practices; for example, a
conversation between two engineers, or between a nurse and a medical doctor. This
connection between the 'general' context and practice is what led Searle to compare his
notion of the Background to Bourdieu's Habitus (Searle, 1992, p. 177).22
A note of caution is in order at this point. Akmajian et al. characterize the three levels of
context described above as "concentric circles". That description, which is represented
schematically in Figure III. 1 below, is intended to indicate the decreasing specificity and
the increasing distance between the utterance itself and the different contexts. However, it
may conjure up a number of images that would be misleading. First, it shows the three
levels as concentric, implying that the outer ones contain the inner ones; that view is
simplistic and seems at odds with the point made earlier that language is the medium of
being. Another problem with this representation is that it shows every circle as having a
crisp boundary, and in particular, it shows the outer one, general knowledge, as clearly
circumscribed. It should be clear from the foregoing discussion that context does not lend
itself to a finite description. Another issue is the implicit distinction that is made between
the linguistic context in the inner circle, and the extra-linguistic contexts outside it. Many
linguists, in particular those who view pragmatics as an all-encompassing perspective, are
opposed to that distinction (Mey, p. 188). That distinction is normal to linguists coming
from the structuralist tradition of transformational generative grammars: it is at odds,
however, with the phenomenological tradition of pragmatics. (It is interesting to note that
an important area of pragmatics has to do with categorization. In opposition to the
rationalist view of categories as being discrete and absolute, pragmaticians have proposed
other models that agree better with findings from psychology and cognitive science.
22
Searle has developed an intricate theoretical framework for explaining intentional phenomena, that is.
phenomena that point to or are about other things. Meanings. understandings. interpretations. but also
beliefs, desires, and experiences, are all intentional phenomena. Searle framework distinguishes between
the "Network" and the "Background". The first comprises the set of other intentional statesthat are
necessary for the intentional phenomenon of interest to function. The "Background", on the other hand,
is a set of capacities that are needed for the network to function, but these capacities cannot themselves
be viewed as more intentional states. Furthermore, Searle distinguishes between "deep Background",
which encompasses those features that are common to all human beings, and "local Backgrounds" or
"local practices", which includes those features that relate to local, cultural practices.
66
These include Wittgenstein's notion of family resemblance, Rosch's concept of a central
prototype, and Lakoff's fuzzy concepts (see Giv6n, op. cit., Chapter 2.))
General knowledge
Immediate physical and
social environment
Linguistic
previousand
anticipated
utterances
The Concentric Circles of Context
Figure 111.
1: Schematic of different levels of context (based on description in Akmajian et al.. 1995)
Dascal (1989) offers a slightly different perspective on context in pragmatics. Instead of
'co-text" and 'context', he prefers 'meta-linguistic' context and 'extra-linguistic' context
('meta' being used here in the sense of 'along with' or 'among'.) As Dascal explains it,
this terminology better reflects the fact that the linguistic context of a sentence or
utterance does not only consist of the text or utterances surrounding it, but also includes
the interlocutors knowledge 'of as well as knowledge 'about' language. Like Akmajian
et al., he also distinguishes three levels of context; thankfully, he does not use their
problematic "concentric circles" metaphor. His view of context can be summarized in a
table, as follows.
67
Table 1112:The diJfrrent kinds and levels of context (based on Dascal, 1989)
Levels of
context 4
Kind of
context -*
Specific or immediate
environment of
production/interpretation
Intermediate (shallow)
General background
knowledge and beliefs
Meta-linguistic
Extra-linguistic
(linguistic)
(situational)
The preceding and following
utterances or text
Who is speaking, to whom,
where, when; the state of
affairs being referred to
Assumptions
about
Assumptions
about the
the
structure of kinds of texts, e.g.,
Assumptions about given set
a poem v. an office memo
activity)23
functioning of verbal
functioning of verbal
communication'
of situations (in Al terms, the
frame or script of a given
Knowledge
about
Beliefs
the about people in
general, their habits and
culture; interlocutor's
about each other
'theories' about each other
"Dynamic" v. "static" context, abduction,and theprovisionalnatureof understanding
Early on in his presentation of pragmatics. Mey cautions against what he calls a "static"
view of context. According to that view. espoused by some linguists, context is "the sum
and result of what has been said up to now. the 'prehistory' of a particular utterance [...]
including the prehistory of the people who utter sentences" (Mey. op. cit.. p. 8). He
explains that it is not enough to take an utterance, look at everything that has come before
it, and predict what it is supposed to r.iean. To illustrate his point, he offers the following
example. Two colleagues are walking together; one asks the other whether he knows the
way to the dining hall, and offers to drive him there. It is only later during the encounter
that they realize that each had understood that the other did not know the way to the
dining hall, and thus needed directions, when in reality they both knew the way. Mey's
point is that at the time of the first ambiguous utterance (Do you know the way to the
dining hall? We can go in my car. ), no amount of a priori contextual information would
have resolved the ambiguity or predicted the illocutionary force of that utterance (that is,
whether it is a question, an invitation. or an offer to exchange directions for a ride.) As
Mey explains:
68
... such a concepton 'context', if establishedindependentlyof the ongoing
interaction between the interlocutors, is completely useless: it is precisely
the dynamic development of the conversation that gives us the clue to
understanding. Such a development cannot be predicted, as it depends
entirely on the individuals and their individual choices at every moment.
The above example highlights one of the pillars of pragmatics, namely, the mode of
knowledge or inference that C.S. Peirce (the founder of pragmatism and "the
acknowledged godfather of modem pragmatics" (Giv6n, p. 20)) called abduction. In
abduction, "one reasons by hypothesis from instances or general rules to their wider
context" (Giv6n, p. 238). "This mode of hypothesis often involves analogical
reasoning, and thus the pragmatic, context-dependent notions of similarity and
relevance" (Giv6n, p. 20).24This mode of inference is particularly interesting for us
because of its obvious relevance to creative activities, of which design and product
development are prime examples.2 ' The first mode of inference, deduction, by which
specific instances are inferred from general rules, and which is emphasized by
rationalists, cannot account for new knowledge. To Wittgenstein, deductive reasoning
(more generally, the propositions of logic) is connected to tautologies or contradictions.
that is, to absolute epistemic certainty (Giv6n, pp. 7, 18, 20, 238-242). The second mode
of inference, induction, involves the discovery of (presumably) general rules from a set of
representative specific cases, and is the mode of choice of empiricists. However, to Peirce
(and, later, Wittgenstein), even this mode of inference cannot account for new knowledge
by itself, because induction involves an element of abduction (Giv6n, pp. 242-244).
Finally, the example from the beginning of this section also highlights an important
aspect of pragmatics, "the fallible and tentative character of every pragmatic
interpretation" (Dascal, 1989, p. 250). This is an aspect pragmatics shares with
23
For a description of frames and scripts, see. for example, Marvin Minsky's "The Society of Mind"
(Minsky, 1988).
24
Peirce's mode of abductive inference is similar to the mode Aristotle called "reduction" or "apagoge", in
which "... the particular facts are not merely brought together but there is a new element added to the
combination by the very act of thought by which they are combined... The pearls are there, but they will
not hang together till someone provides the string..." (Aristotle, Posterior Analytics, as quoted in Giv6n,
1989, p. 14.)
25 Giv6n also
describes it as proceeding "by hypothesis, guesswork or intuition, often by analogy." "It is
thus, in principle, unconstrained" (Giv6n, 1989, p. 7).
69
hermeneutics. another phenomenological theory of understanding and knowing that is the
subject of the next section.
Background II: Hermeneutics
Beside linguistics, the framework that is presented later in this chapter draws on ideas
from the field of hermeneutics. Hermeneutics is defined as "the art or theory of
interpretation, as well as a type of philosophy that starts with questions of interpretation"
(Audi, 1995, p. 323). Hermeneutics was initially concerned with the interpretation of
sacred texts, but over time. this narrow concern broadened considerably. Hermencutics
now refers to an important position of German philosophy in this century, a position that
has influenced the French philosophers of the last several decades (e.g., Riceur and
Merleau-Ponty.)
The following description of hermeneutics, from David Linge's introduction to
Gadamer's "Philosophical Hermeneutics" hints at the potential usefulness of
hermeneutics to the PD problems that are the concern of this research (Gadamer, 1976):
Hermeneuics has its origin in breachesin intersubjectivity.Its field of
application is comprised of all those situations in which we encounter
meaningsthat are not immediatelyunderstandablebut requireinterpretive
effort. The earliest situations in which principles of interpretation were
worked out were encounters with religious texts whose meanings were
obscure or whose import was no longer acceptable unless they could be
harmonized with the tenets of'theifaith. But this alienation from meaning
can just as well occur while engaging in direct conversation, experiencing
a work of art, or considering historical actions. In all these cases. the
hermeneutical has to do with bridging the gap between the.familiar world
in which we stand and the strange meaningthat resists assimilationinto
the horizonsof our world
This section is a brief look at hermeneutics that focuses on some of the key concepts that
will be used in the classification framework that follows. This section draws heavily on
Palmer's presentation of the subject (Palmer, 1969).
The word hermeneutics is derived from the Greek term for interpretation (the verb
hermeneuein, which is translated as 'to make clear', or 'to interpret'), which is itself
70
related to the Greek god Hermes. In Greek mythology, Hermes, who, interestingly, is
credited with discovering language and writing, was the messenger and herald of the
other gods, "associated with transmuting what is beyond human understanding into a
form that human intelligence can grasp." These roots of the word hermeneutics suggest
the process of"bringing to understanding" in particular through the medium of language
(Palmer, op. cit., p. 13).
The three meanings of hermeneutics
In its ancient usage, 'hermeneuein'
had associated with it three related meanings. or
"three basic directions of meaning." These are 'to say', 'to explain', and 'to translate'.
The following quick look at these three 'directions of meaning' of hermeneutics will
introduce the basic concepts of hermeneutics, and the issues that the field is concerned
with.
Interpreting as 'saying'
The first meaning, 'saying' or 'expressing aloud in words', suggests that an important
aspect of interpreting something involves simply saying it or asserting it. That sense of
interpreting is intuitively obvious in the case of a performance, be it a reading, a play. or
a musical performance, and it contains within it the notion of 'style', the way of
expressing that is particular to the performers. This sense of interpreting as 'saying'
already points a few key aspects of hermeneutics. First, there is the notion that
interpretation as a creative act. Palmer explains this using the case of oral interpretation.
or reading aloud, which he compares to a musical performance:
Considerthe act or reading aloud.Oral interpretationis not a passive
responseto the signs on the paper like a phonographplaying a record;it
is a creativematter, a performance.like that of a pianist interpretinga
piece of music.Any pianist can tellyou that a musicalscore itself is a
mere shell; the "meaning" of the phrases must he grasped to interpret the
music.
Another important notion in hermeneutics is the circular character of coming to
understanding. Again, using the case of oral interpretation, it is clear that one must grasp
71
the meaning of what one is about to express in order to be able to express it properly,
with the proper emphasis and intonation. As Palmer explains:
Oral interpretation thus has two sides: it is necessary to understand
somethingin orderto expressit, yet understandingitselfcomesfrom an
interpretivereading-expression.
Finally, interpreting as 'saying' provides another perspective on the limitations of
information transfer/information processing view of communication that were discussed
in the introduction. According to Palmer, understanding information is not the same as
communicating a message; it is a different use of language:
"Inbfrmation " is a significant word [...]. It appeals to the rational ficulty
and not to the whole personality; we do not have to call upon our personal
experienceor risk ourselvesin orderto understandinfbrmation--and
infrmation does not suffer much from silent reading.
Interpreting as 'explaining'
Whereas the first sense of interpretation focused on its expressive aspect, this second
sense places the emphasis on the explanatory dimension of interpretation, on the fact that
words are generally used to say something about something else. be it to explain it or
clarify it or rationalize it. If 'saying' or 'expressing' can be thought of as a first moment
in interpretation, then 'explaining' would be "a second moment of interpretation."
Explaining emphasizes the discursive aspect of understanding. an aspect that draws on
reasoning and the intellect. not intuition. However, Palmer cautions against thinking of
this second moment of interpretation in terms of logical analysis. In making that point, he
relies on Aristotle's definition of interpretation as 'enunciation'. "the primary operation
of the intellect in formulating a true judgment about a thing" (Palmer, op. cit., p. 21 ).26 As
such, interpretation falls into a particular category of Aristotle's operations of the
intellect, one that comprises operations of composing and dividing, a category that is
distinct from the one that comprises opei tions of reasoning from known things to
unknown things. that is, the category of logical analysis.
26
This is somewhat counterintuitive, as 'enunciation' seems closer in meaning to 'saying' and 'expressing'.
72
This distinction hints at a number of related and important concepts in hermeneutics, such
as Heidegger's rejection of intentionality and the distinction between subject and object
that it presupposes (Dreyfus, 1991, pp. 46-54). More directly, it shows up in his
description of the different ways that one encounters objects and tools, in which the
analytical stance becomes operative in situations where the transparency of equipment is
lost as a result of a breakdown (Dreyfus, op. cit., p. 70).
Although logical analysis is itself a form of interpretation, it is a derivative form, since
logic works by comparing already enunciated statements. The idea that a prior or
preliminary interpretation serves as the basis for another interpretation points out another
notion in hermeneutics, related to the problem of the subject/object dichotomy. It is the
rejection of the idea that method and object can be separated, which Gadamer addresses
in Truth and Method. Palmer explains the issue thus:
... method has already delimited what we shall see. It has told us what the
objectis as object. For this reason,all methodis alreadyinterpretation.
This idea of one level of interpretation framing another level brings us to a very
important issue in hermeneutics, one that it has in common with pragmatics, namely. the
role of context in understanding. Given the rejection of the separation between subject
and object, then it is not possible for an object to have any significance in itself, "outside
a relationship to someone", that is, apart from a perceiving subject. Similarly, an
explanation is always for someone; 'all explanatory interpretation assumes intentions in
those to whom the explanation is directed" (Palmer, p. 24). This context of already given
meanings and intentions, that is, of 'assumed understanding", is referred to in
hermeneutics as 'preunderstanding', or the 'horizon' of the person seeking the
explanatory interpretation.
In hermeneutics, explanatory interpretation, or coming to understanding in general,
involves a merging of horizons: the horizon 'inhabited' by the literary work, or the work
of art, or the interlocutor, on the one hand: and the horizon of the interpreter, that is. his
"world of intentions, hopes, and preinterpretations", on the other.
73
Finally, the notion of preunderstanding provides another reminder of the 'hermeneutical
circle', through which one comes to understanding: some understanding is always
necessary for further understanding.
Interpretation as 'translation'
This third 'direction of meaning' of hermeneutics-interpretation as translation-is a
special case of the process of bringing to understanding. Here, the contrast or clash
between the horizon of the interpreter and the horizon of that which is being interpreted is
readily obvious. At a basic level, this could result from the work being written in a
foreign language. Since language itself is "a repository of cultural experience", the
interpreter as translator has to bring what is initially foreign and unintelligible into the
world of his or her own language. The process of translation is therefore a process of
mediation.
The clash of horizons or perspectives that makes interpretation more like translation does
not only occur when there is language difference. It could be the result of large difference
in time, as when interpreting ancient texts. In this case, the process of mediation does not
only take place through language, but also through history. In this case, "an effort of
historical imagination" is required to envisage the world of the work (Palmer, p. 31).
The notion of translation is relevant even when there is no difference in language nor
time. but where the conflict of horizons is the result of very different worldviews. as
might exist between scientists and marketers for example. The challenge in explaining
scientific issues to marketers, or vice versa, remains one of bringing what is "strange,
unfamiliar, and obscure in its meaning into something meaningful that speaks 'our
language."' This involves an explication of the 'other' worldview, and a explication of
its significance to us.
Theevolutionof the field
Palmer identifies six modem definitions of hermeneutics that are connected to the
evolution of the field, at least since 1654. These six definitions are not only different
stages of historical development; they also reflect different approaches to the problem of
interpretation.
74
The first definition, hermeneutics as a theory of biblical exegesis, reflects the original
concern of the field with developing a body of rules and methods, a system, for
interpreting Scripture. In the second definition, that of herrneneutics as a methodology for
philology, the focus remains on rules and methods of exegesis, but the scope of the field
is broadened from the strictly religious to any text. Since it coincided with the
development of rationalism, this stage is characterized by an effort to eliminate advance
judgments in interpretation. Interpreters focused on looking beyond the (inferior)
"accidental truths of history", and finding in the text the (superior) "truths of reason" that
would be "relevant to the enlightened rational man." The task of the interpreter thus
became a historical one, as only full knowledge of the historical context of the work
would make it possible for the interpreter to purge the historical elements in it and grasp
the spirit behind it.
The third definition is of hermeneutics as the science of linguistic understanding. a phase
associated with the philosopher Schleiermacher. liHeis credited with creating a modern
general hermeneutics, a "science" or "art" of understanding in general. Instead of an
assemblage of rules for philology, he sought a hermeneutics that would describe the
conditions for understanding in all forms of dialogue. Following Schleiermacher. Dilthey
sought to make hermeneutics the methodological foundation for the
Geisteswifenschaften.
that is. the arts, humanities, and other disciplines that focus on
man's actions. This fourth definition of hermeneutics was again focused on interpretation
as requiring an act of historical understanding. an act that drew upon the interpreter's
personal knowledge of what it is to be human.
The fifth definition listed by Palmer is "hermeneutics as the phenomenology of Dasein
and of existential understanding." This definition of hermeneutics is associated with
Heidegger and, later, with Gadamer. Unlike the previous approaches to hermeneutics.
which were epistemological in nature. Heidegger was concerned with the ontological
problem of man's 'everyday being-in-the-world' (Dasein. which literally translates to
"being there", and colloquially means "everyday human existence", is used by Heidegger
in a special way. He uses it to denote 'human being', i.e., man's 'way of being'. as well
as 'a human being', but not in the Cartesian way in which the individual plays the central
75
role of conscious, "meaning-giving transcendental subject" (Dreyfus, 1991. pp. 13-14).)
In Heidegger's analysis, interpretation and understanding are fundamental ways of man's
being. Dasein's way of being, that is, the way of being that is shared by human beings,
institutions and cultures, is unique because it "embodies an understanding of what it is to
be", i.e., the capacity for self-interpreting (Dreyfus, p. 15). It is his notion of this
ontological significance of understanding that makes Heidegger's contribution to
hermeneutics revolutionary (Palmer, p. 42).
Gadamer developed and refined Heidegger's system, focusing on the role of language in
being, and the linguistic character of human reality (ibid.).
This fifth definition of hermeneutics is the one most relevant for the purpose of
developing the classification framework of PD practices. I will return to it in greater
detail in the next section.
Finally, the sixth definition of hermeneutics. according to Palmer, is as a system for the
interpretation of signs and recovering their hidden meanings. This marks a return to the
role of hermeneutics in exegesis, except that in this case. the 'text' can be any symbol,
myth or dream. This definition, most strongly associated with Ricoeur. is not directly
relevant to our interests here.
The brief description above of the different modern definitions of hermeneutics not only
shows the shift in the focus of hermeneutics over time: it also highlights the significant
differences in the philosophical concepts that underlie these different definitions.
Generally, this difference can be summed up in terms of the two competing positions in
modern hermeneutics. According to the first position. represented by Dilthey.
interpretation is a method for the human sciences, focusing on rules and criteria for
extracting from the text the original meaning its distant author intended. This is similar to
the notion of speaker's meaning that we saw earlier in linguistics. The second position.
following Heidegger. views interpretation as an 'ontological event,' an interaction
between interpreter and text that is part of the history of what is understood.'
An example
of this notion of interpretation is what happens when a court or a judge interprets a law in
the context of a case: through that process of interpretation, the law is itself invariably
transformed (Audi, p. 323).
76
So far, this overview of hermeneutics has been rather general. Before we can use
hermeneutics in the theoretical framework to follow, we need to take a closer look at
some of the key concepts of the field. This is the objective of the next section.
Key elements of hermeneutic interpretation
This section expands on some of the key aspects of hermeneutics that were briefly
described in the previous sections. The focus here is on Heidegger's and Gadamer's
'philosophical hermeneutics'. This presentation is, by necessity, partial and biased
towards the goal of this chapter.
Before proceeding, another caveat is in order. The concepts and principles of
hermeneutics, based as they are on Heidegger's conceptions of thinking, human
experience, and history, may seem counterintuitive, despite the fact that they were
intended to better describe and reflect real-life human experience. The reason is our
tendency to lapse back into the more familiar scientific cognitive conceptions of
understanding and interpretation. Another source of difficulty is that these concepts are
interrelated, and it is difficult to present them in a clear, linear, hierarchical way, or, more
generally, as a set of mutually exclusive, collectively exhaustive set of principles. Palmer
cautions that "we can only gradually enter into the circle of his [Gadamer's]
consideration." And that "the obstacles to understanding Gadamer's thought are
formidable" (Palmer, p. 166).
It might be worthwhile, before proceeding, to summarize what Gadamer's hermeneutics
is about. Gadamer's major contribution was transforming hermeneutics "from a subdiscipline of the humanities concerned with the techniques of interpretation of theological
and classical texts ('classical' hermeneutics) and from the distinctive methodology of the
Geisteswifienschaften ('romantic' hermeneutics) into a reflection on the fundamental and
universal conditions of experience" (Dascal, 1989). To him, understanding "is always an
historical, dialectical, linguistic event." whether in the humanities, in the sciences, or in
any other arena of human endeavor (Palmer, p. 215). Understanding is no longer
conceived as the act of a human subjectivity against an object, as in the traditional
Cartesian cQnception. And Gadamer calls into question the status of method as the way to
truth. Methodical forms of consciousness, be they aesthetic, historical, hermeneutic (in
77
the classical, narrow sense), or scientific can only capture parts of our experience in the
face of the corresponding 'objects' (works of art, historical events, texts, scientific
experiments.) Method invariably involves a certain reductionism and therefore cannot
account for those aspects of experience that fall outside the scope of the method (Dascal,
op. cit.).
This section is organized following Dascal's approach in his paper comparing pragmatic
and hermeneutic interpretation (Dascal, op. cit.). Certain elements of Gadamer's
philosophical hermeneutics are presented as contextual factors that play an important role
in interpretation; others are presented as crucial characteristics of the process of
interpretation itself.
Contextual factors in hermeneutic interpretation
Dascal identifies three contextual factors that play a role in understanding: a) the question
that implicitly motivates every statement, utterance, text, or more generally, work; b) the
prejudgments or horizon of the interpreter; and c) the role of language as the universal
medium of all human experience, of the 'heritage'. The present discussion will focus on
the first two factors (Dascal, op. cit., p.243).
The motivating question
To Gadamer. every experience has a dialectic character, by virtue of its encounter with
negativity. Palmer quotes the following from Truth and Method (Palmer. p. 19):
Experience is a matter of multisided disillusionment based on expectation:
only in this way is experience acquired
And:
Every experience runs counter to expectation if it really deserves the name
experience.
The connection between this encounter with negativity and questioning is clear:
... in all experience, the structure of questioning is presupposed The
realizationthatsome matter is other than one hadfirst thought
presupposes the process ofpassing through questioning. (From Truth and
Method, as quoted in Palmer, p. 198.)
78
The central importance of questioning is clear in the following quote from Gadamer's
essay entitled The Universality of the Hermeneutical Problem, describing the
Urphanomen (the original, primal phenomenon) of hermeneutics (Gadamer, 1976, p. 11):
No assertionis possiblewhich cannot be understoodas an answer to a
question,and assertionscan only be understoodin this way.
Dascal understands that to mean that being "somehow aware" of the question that
motivates a statement is necessary for understanding that statement, thus giving the
motivating question a context-like role. In that way, the role of 'the question' in
hermeneutics is similar to that of 'conversational demand' (an element of co-text) and
'pragmatic appropriateness' in pragmatics (Dascal, op. cit., p. 244).
The concept of 'the question', and its contextual role, is not restricted to utterances and
conversations. In view of the dialectical character of all understanding, it can apply
equally to a whole text, a theory, or any non-linguistic action. But Dascal points out that
as one moves away from the "relatively clear-cut case of an actual conversation", the
concept of 'the question' becomes vague and difficult to identify, finally becoming
subject to interpretation itself, instead of being a reference to be used in interpreting:
Furthermore,rather than a contextualgiven that can be relied uponfor
the interpretationof a problematicutterance, 'thequestion' becomesitself
something that, like the interpretation one is seeking. must be inferred
from the utterance and other contextual information, and thereby loses
much of its usefulness as an aid to interpretation (Dascal, op. cit., p. 245).
This criticism is somewhat misplaced, on the following two (related) counts. First, it
doesn't take into account the circular nature of understanding (although Dascal does
discuss the hermeneutic circle later on, as one of the characteristics of the hermeneutic
interpretation process.) Secondly, the statement discloses a highly instrumental view of
'the question' as a contextual object to be used; that view is at odds with Heidegger's and
Gadamer's thinking.
Because of his focus on 'the question' as a contextual object and, more generally, his
intention to compare the notion of context in hermeneutics to that of pragmatics, Dascal
does not discuss some of the more interet3ng aspects of the question and the dialectics of
79
understanding. These relate to Gadamer's conception of what genuine questioning is.
Starting with a quote from Wahrheit und Methode, Palmer describes the characteristics of
genuine questioning as follows:
"In order to be able to question one must will to know, and that means,
however, to know that you do not know. " When one knows he does not
know, and when he does not therefre through method assume that he only
needs to understand more thoroughly in the way he already understands,
then he acquires that structure ofopenness characterizing authentic
questioning. [...] The openness of questioning, however, is not absolute,
because a question always has a certain direction. The sense of the
question already contains the direction in which the answer to that
question must come, f it is to be meaningful and appropriate (Palmer, op.
cit., pp. 198-9).
Genuine questioning, therefore, presupposes a certain openness. namely, the desire to
know, possibly in a different way; it also involves specifying certain boundaries. Genuine
questioning involves a dialogue between the interpreter and the text, or the interpreter and
the work of art, or, in our case, the interpreter and his or her interlocutor. Palmer
describes a dialogue as the opposite of an argument: whereas in the latter, one tries to
defend and hold on to the opening point or response, a dialogue requires openness and
readiness to move from the initial position into unforeseen directions. Palmer quotes from
Truth and Method (Palmer, op. cit., p. 199):
A dialogue does not try to argue down the other person, but one tests his
assertions in the light of the subject itself:
This point hints at another element that is critical to 'authentic questioning'. namely. the
total immersion by the interpreter in the subject matter. "According to Gadamer. there is
only one way to find to find the right question. and that is through immersion in the
subject itself' (Palmer, ibid.). This total immersion in the subject matter, as opposed to
the traditional Cartesian stance of a subject querying an object. reminds us of an
important hermeneutic principle. It also reminds us of Gadamer's concept of the game.
which he uses to describe the total immersion necessary for understanding. The following
is from Linge's introduction to Philosophical Hermeneutics:
80
This element of buoyancy-of being borne along by the subject matter-is
illuminatedby a secondphenomenonthat Gadamerdescribesin support
of'this theory of understanding, the phenomenon f the game or playing.
Even more strongly than the analogy of the conversation, Gadamer 's
phenomenology of the game suggests the inadequacy of'trying to
comprehendunderstandingfromtheperspective of the subjectivityof the
authoror the interpreter.[...] For what reveals itself as most
characteristicqf the phenomenonof playing is that the individualplayer is
absorbed into the back-and-forth movement of'thegame, [...] and does not
hold back in sel-awareness as one who is "merely playing." Theperson
who cannot lose himself inJll earnest in the game or give himself over to
the spirit of/thegame, but instead standsoutside it, is a "spoilsport " one
who cannot play (Gadamer, 1976, pp. xxii-xxiii).
Given the importance of genuine questioning to hermeneutics, it would have been equally
appropriate to deal with it later in this section, as one of the characteristics of the process
of interpretation, instead of presenting it as a contextual factor.
Prejudgment or 'prejudice'
Schleirmacher and Dilthey, whose hermeneutic position was mentioned earlier, identified
the meaning of a text or an action with the subjective intention of its author. In this view,
"understanding is essentially a self-transportation or imaginative projection whereby the
knower negates the ... ] distance that separates him from his object [...]" (Gadamer, op.
cit., p. xiv). Meaning can only be achieved through a careful and thorough reconstruction,
and full awareness of the original life-context or life-world of the author. The knower's
own present situation is "a source of prejudices and distortions that block valid
understanding" (ibid.).
Although they were concerned with historical texts, one can see the similarity to a
contemporary line of thinking in product development, which emphasizes the importance
of 'becoming one with the customer', or which asks marketers to 'think like engineers'
and vice versa.
Gadamer's objection to the 'romantic' hermeneutics of his 19th century predecessors
centered on their neglect of a critical dimension in understanding (according to his own
hermeneutics), namely, the reflexive dimension. The following is from Truth and
Method:
81
The self-interpretation [Selbtbesinnungf of the individual is only aflicker
in the closed stream of historical life. For this reason the prejudgments Qf
the individualare more than merelyhisjudgments; theyare the historical
reality of his being (Palmer,op. cit.,p. 182).27
Therefore, asking the knower to alienate himself from his own historicity is asking for the
impossible, since one's historicity is not accidental (nonessential) but rather ontological.
that is, constitutive of one's being. The result is that there can be no "presuppositionless"
interpretation or understanding. The fact that we are tied to our present horizons, and the
distance that separate us as knowers or interpreters from our object of interpretation, are
in fact "the productive ground of all understanding rather than negative factors or
impediments to be overcome" (Gadamer. op. it., p. xiv). In talking about our historicity
and finitude, Gadamer uses the term "prejudice". not in the negative sense of an
unfavorable opinion or feeling, but in the "literal sense":
It is not so much ourjudgments as it is our prejudices that constitute our
being. This is a provocativeformulation,for I am using it to restoreto its
rightfulplace a positive concept ofprejudice that was driven out ofour
linguistic usage by the French and the English Enlightenment. [...]
Prejudicesare not always unjustriedand erroneous...(Gadamer,op. cit.,
p. 9).
He goes on to describe the productive role of prejudices:
Infact, the historicityof our existenceentailsthat prejudices,in the literal
sense of the word, constitute the initial directedness ?f our whole ability to
experience.Prejudicesare biases of our opennessto the world. Theyare
simply conditions whereby wseexperience something-whereby what we
encounter says something to us (ihid.).
The issue of prejudice is summarized by Dascal as follows: first. the interpreter cannot
free himself from the multitude of prejudgments and categories which are constitutive of
his being; secondly. interpretation is impossible if the interpreter is a "tabula rasa",
devoid of conceptual categories, values, and prejudgments (Dascal, op. cit.., p. 245).
27 Gadamer
apparently uses "Besinnung"
to mean reflection (on). The more common meaning of
Besinnung is consciousness.
82
Before closing this section, it is worth noting the conlnection Dascal makes between the
interpreter's prejudgments as a contextual factor in hermeneutic interpretation, and the
typology of contexts presented in the section on pragmatics. Prejudgments would seem to
fall most naturally within the category of 'extra-linguistic background'. He points out
however that "in light of Gadamer's [effort] to generalize and extend both notions [the
'question' and 'prejudice'], each of them can finally come to apply to any of the levels
and kinds of context discussed above" (Dascal, op. cit., p. 248).
As was the case with the 'question', focusing on 'prejudice' as a contextual factor is not
the most productive use of the concept. The power of the notion of prejudice lies in its
implications for the role of the interpreter (which Dascal does examine.) These are
discussed in the next section.
Prejudice and the constitutive role of the interpreter
One of the most central, yet most difficult, problems in hermeneutics is the problem of
meaning of the text or of the work, and the challenge of fidelity to the meaning. In
Linge's words, "The customary way of defining the meaning of a text has been to
identify it with the subjective act of intending of its author." On that view, the problem of
understanding becomes one of recovering the original intention of the author. That view
also makes it possible to talk about a "canonical interpretation, the one that captures 'the
correct [meaning] and banishes all competing interpretations as incorrect" (Gadamer, op.
cit., p. xxiv). This notion of meaning requires that we distinguish between the various
explications of the significance a text may have to us on the one hand, and the one unique
meaning that its creator had in mind, and that he put in the text on the other. This
distinction does not hold against the test of history: over time, people have not only
differed on the significance of a text to their lives; they have also had fundamental
disagreements "on what they thought they saw in the text..." (ibid.). Gadamer therefore
rejects this traditional view that restricts the meaning to the mens auctoris (the author's
mind.)
The meaning of a lext surpasses its author not occasionally. but always.
Thus understanding is not a reproductive procedure, but rather always a
productive one... It suffices o say that one understands differently when
83
one understands at all. (Quoted by Linge from Wahrheit und Methode, in
the introduction to Gadamer, op. cit., p. xxv.)
Understanding therefore has a creative element, according to Gadamer. This is counter to
the traditional view which, by confining meaning in the mens auctoris, reduces
understanding to a "transaction between the creative consciousness of the author and the
purely reproductive consciousness of the interpreter" (Gadamer. op. cit., p. xxiv).
According to Dascal. this aspect of hermeneutic interpretation represents the most
significant difference between hermeneutics and pragmatics (Dascal. op. cit.). In both,
incorporating the relevant contextual factors is key to achieve understanding. In
pragmatics the ftcus is on the 'context ot'production',. that is. on "the historicity and
situation-boundedness
of the author. speaker. or text..." (ibid.). However, by highlighting
the positive role of prejudice in creating meaning and coming to understanding.
hermeneutics eff'ectively shifts the emphasis to the 'context of interpretation'. In other
words, the focus now is on the historicity and situation-boundedness
of the interpreter or
addressee. This shift has clear and important implications concerning the role of the
interpreter:
The picture that emerges. rom this .shifiis that of an inierpreter iwhois not
(the contextual inormatlion that
only active in the sense o/ gathering tll
contributes to understanding the meaning that is in the text (presumahly
byhhaving been put there by the author), hut. more undamentalll, in the
sense of somone who creates meaning and is constitutive f meaning
through his very activity qua interpreter (Dascal. op. cit.. p. 246).2)
In the case of music and drama, the contribution of the interpreter to the generation of
meaning is even more significant. as a result of what Gadarner calls "the excess of
meaning that is present in the work itself' (Gadamer. op. cit.. p. 102).
Dascal's characterization of this shift in emphasis may reflect the significance of the
impact he foresees hermeneutics having on his field of pragmatics. However. it probably
overstates the emphasis of hermeneutics itself It is not that hermeneutics shifts the
emphasis from the 'context of production' to the 'context of interpretation'. as much as it
28 Qua: Latin adverb meaning "as". "as being". "in the character or capacity or'.
84
shows that the historicity and situation-boundedness of both author and interpreter play
equally important roles in bringing creating meaning. This can be clearly seen in the
concept of the 'fusion of horizons', one of the aspects of hermeneutic interpretation that
are discussed later in this section.
Some important aspects of hermeneutic interpretation
As was clear in the previous section on contextual factors, it is difficult to separate these
factors from the process of interpretation itself. Similarly, it is difficult to separate the
different aspects of hermeneutic interpretation from one another and discuss them
separately. Nevertheless, Dascal identifies four 'crucial aspects of the process of
interpretation" as conceived by Gadamer. These are the 'circular' and provisional nature
of understanding; the fusion of horizons that takes place when interpretation is successful
and the interlocutors come to understanding; Gadamer's concept of 'effective-historical
consciousness' or 'authentically historical consciousness' and its role in interpretation;
and, finally, tradition and its role.2 9
The notion of the hermeneutic circle deals with the difficulty one faces in coming to
understanding. as a result of the fundamental relationship between the whole and the
component parts of the thing to be understood. This difficulty is aptly captured in the
following quote from Friedrich Ast, a philologist concerned with old texts from Greek
and Roman antiquity, and to whom capturing the spirit or Geist of the period was
essential to understanding its ideas and feelings (Palmer, op. cit., pp. 76-77):
One can only rightly grasp the combined unity of the spirit of antiquity if
one graspsthe individualrevelationsof it in individualancient works,and
on the other hand, the Geist of an individualauthorcannot be grasped
apartfrom placing in its higher relationship [to the whole]. (As quoted in
Palmer, p. 77.)
The hermeneutic circle operates at all levels of understanding. The previous quote deals
with a case where the whole happens to be the spirit of an age, but the common
29
The first expression is Dascal's translation; the second is Palmer's (Dascal. op. cit., p. 249; Palmer, op.
cit., p. 191). Gadamer's original expression is wirkungsgeschichiliche Bewu,&sein. which can be
translated literally as "'consciousness in which history is ever at work,' or 'historically operative
consciousness."' (See Palmer, ibid.)
85
description of the circle operates at the level of words, the whole being the sentence
formed by these words. The sentence cannot he understood unless one grasps the
meaning of the individual words; yet the words derive their meaning from the immediate
context in which they find themselves, the sentence itself.
Clearly, the concept of the hermeneutic circle has a built in logical contradiction: "if we
must grasp the whole before we can understand the parts, then we shall never understand
anything" (Palmer, op. cit., p. 87). This is the reason why the hermeneutic circle appears
like a vicious or impenetrable circle. One answer is that logic cannot fully explain how
this process works. To Schleiermacher. tor example. understanding involves a
comparative dimension (where we compare the new to something that we already know).
as well as an intuitive. With these dimensions. it is possible to think of a "leap" into the
hermeneutic circle whereby we understand the parts and the whole together (ibid.).
In Gadamer's hermeneutics however. this logical contradiction is a non-issue. Speaking
of whole and parts already assumes a subjective process involving man over and against a
number of objects (viz.. the whole and its parts.) But as we saw earlier. Gadamer's
philosophical hermeneutics (following Heidegger) rejects such a notion of understanding,
seeing understanding as man's main ontological process. Dascal explains it as follows:
Given the constitutive role of the interpreter s 'prejudgments. ' which
implies that there is no such thing as 'the meaning 'of a text to he
'discovered, 'coupled with the frict that such pre-judgments are themselves
'of the
not fixed. but shifting (among other things due to the 'resistance
'thingitself' expressedin the text) in the courseof the processof
understandingso that the interpreterhimself is shapedbythe process the
hermeneuticcircle must be understoodas an ontologicallybasic level
prior to the distinction between subject and object (Dascal, op. cit.. p.
249)
According to Dascal, this "radicalization" of the concept of the hermeneutic circle, that
is, ascribing to it a "positive ontological role," is what makes it finally possible to resolve
the seeming contradiction associated with it:
... the circle is. in fact, a siral of successive. more encompassing
understandings. These are achieved not by 'neutralizing' the interpreter :v
horizon of pre-judgments 'nor by ignoring the text s contextual horizon,
86
but by taking full advantage of'the interplay between both, leadingto a
'fusion (ofhorizons' (Dascal, op. cit.. p. 250()).
This process is one where "partial understanding is used to understand still further..."
(Palmer, op. cit., p. 25). Whatever understanding is achieved at a given point is subject be
revision as the 'knower' moves outwards along the spiral. Understanding is therefore
speculative and always provisional, "likely to be revised or entirely replaced as in light of
contextual factors previously overlooked" (Dascal, op. cit.. p. 249). The 'knower' starts
from his present situation, a situation that does not hold a privileged status. as was the
case with the prejudice-ree
position of the older, "scientific" hermeneutics. Rather, this
present situation is "a fluid and relative moment in the life of effective history..." (Linge.
in Gadamer, op. cit., p. xix). This situation, or moment. is disclosive, since it provides the
pre-understanding of the subject matter that makes communication at all possible: it is
also productive, because of the constitutive role prejudice plays in the event of
understanding. But that moment is soon tested through the encounter with the subject
matter. and, "like all others before it, will be overcome and fused with other horizons"
(ibid.). The process of fusion of horizon is illustrated by CGadamer'sexample of a
successful discussion (as quoted in Dascal, op. cit., p. 251):
We are conlinually shaping a common perspecti!hewhen we speak a
common language and so are active participant in the communality of
our experienceof the world.Experiences(f resistanceor oppositionhear
witnessto his,.fir example.in discussion.Discussionbearsfruit when a
common language is found Then the participants part from each other as
changed beings. The individual perspectives with which they entered upon
have been tran.sformed,and so they,are transformed themselves. This.
then, is a kind of progress-- not the progress proper to research in regard
to which one cannot ill behind. but a progress that alwas must be
renewed in the effort of our living (Gadamei, 1981).
This quote underscores, again, the ontological nature of understanding in Gadamer's
philosophical hermeneutics. But in order fully to understand it, we need to talk about the
role of language according to Gadamer (and Heidegger before him), something that was
touched upon early in this chapter. Our prejudices. which. on Gadamer's account, are
critical to our ability to understand. are embedded in and passed on through our language.
Or, as Linge explains: "Since our horizons are given to us prereflectively in our language.
87
we always possess our world linguistically" (Gadamer, 1976, p. xxviii). And later, he
talks about Gadamer's agreeing with Heidegger "that language and understandingare
inseparable structural aspects of human being-in-the-world, not simply optional functions
that man engages in or does not engage in at will" (ibid., p. xxix). Understanding and its
conception as involving a fusion of horizons is therefore "an essentially linguistic
process."
The above description of the spiral progression of understanding highlights the dialectical
nature of the process (an aspect that was discussed earlier in the context of 'the
question'.) In this, Gadamer is compared to Hcgel. "who saw that knowledge is a
dialectical process in which both the apprehending consciousness and its objects are
altered" (Gadamer, 1976. p. xxxix (I,inge's introduction)). Ilegel's view of new
knowledge as "a mediation or refocusing of the past within a new and expanded context."
and Giadamer's conception of understanding as a fusion of horizons both share the same
"dynamic and self-transcending character" (ibid., pp. xxxix-xl). The result is an ever-
increasing consciousness. "
"Effectivehistoo "
This consciousness that is ever expanding is "conceived by Gadamer as a self-critical
awareness of the historicity--hence finitude but openness. limitation but changeabilityof the different but not entirely alien horizons involved in understanding" ( I)ascal. 1989.
p. 250). Gadamer's concept of 'effective history'. or of the 'historical-effective
consciousness', is critical to his hermeneutics. But how useful or necessary is it or our
purpose here?
There
T"
are at least two critical differences between Gadamer's and 1Hegel'sdialectics. One is that Hlegel
starts from a subjectivist
viewpoint,
that is. he takes "the human subjective consciousness and the
certainties of reason based on it. as the ultimate point of reference for human knowledge" (Palmer. 1969,
p. 164). In other words.
whether something is seen as true or false depends on the beliefs, wants and
hopes of the 'knower'. Gadamer. following Heidegger. rejects the subject-object dichotomy, and views
thinking as part of being, where one lets himself be guided by the thing being understood itself. Palmer's
insistence on the subjectivism
of Hegel seems strange. as Hegel had argued against the priority
of the
individual, self-conscious subject. His notion of Geist, or human 'spirit' is indeed in opposition to the
individual subject; still, it can be thought of as a 'collective subject' (Audi, 1995, p. 3 14). The second
important difference between Gadamer's and Hegel's dialectics lies in the latter's belief in an 'Absolute
Spirit', that is, the possibility of the full and complete of human self-knowing (ibid.). Gadamer. on the
other hand, the 'elevation
to a higher universality'
one reaches through understanding.
88
and through
First, we need to understand the concept of 'effective history' somewhat better. Gadamer
intends;' to stress the temporal character of interpretation; one of his main concerns was
to make clear that "[U/nderstanding is not reconstruction but mediation," and a
"'genuinely productive" experience (ibid.. p. xvi). In other words. 'knowers' and
interpreters are "conveyors of the past into the present" (ibid.). Gadamer emphasized "the
fundamental continuity of history as a medium encompassing every such subjective act
[of understanding] and the objects it apprehends. Understanding is an event, a movement
of history itself in which neither interpreter nor text can be thought of as an autonomous
part" (ibid.). Therefore, against the notion of the past as "a collection of objects to be
recovered or duplicated by the interpreter" Gadamer proposes his concept of the past as
an 'effective history'. the medium that "alone makes possible the conversation between
each new interpreter and the text or event he seeks to understand" (ibid.. p. xvii).
It is tempting to see Gadamer's focus on history as a reflection of his initial concern with
old texts, and with bridging the temporal gap between these texts and modem
interpreters. In that case, the concept of 'eft'tective history' would have limited usefulness
in the framework proposed here. But 'effective history' is critical to Gadamer's concept
of 'tradition', the most important contributor to the 'commonality of understanding'
which is one of the conditions for entering the hermeneutic circle. Dascal explains this
relationship as follows:
[Historical-f/ictive cons.ciousness does not purport to free usfirom the
hound of 'tradition '---an impossible task ---hut rather to make us Laware
of
it and. thus. to put us in aIposition to ascend the spiral of understanding
through a full exploitation of the 'hermeneutical experience' (Dascal.
1979,p. 250).
Perhaps the best way to utilize the concept of historical-effective consciousness in the
framework that follows is to return to Gadamer's typology of the 'I-thou' relationship.
which he uses "to help situate and thus clarify the nature of the historically operative
consciousness (Palmer, 1969. p. 191 ). Gadamer uses three typologies: a) thou as an
object within a field. b) thou as reflexive projection, and c) thou as tradition speaking.
following the spiral outwards. "remains finite and surpassable and is not to be equated with Hegel's
absolute knowledge in concepts." (See the introduction by Linge. in Gadamer. 1976, p. xl.)
89
The first I-thou relationship is that of scientific objectivity and methodological knowing.
The thou is seen in an instrumental way, as something specific. an object within one's
field of experience that can be used as a means to achieve one's objectives. In this case,
the thou is understood in terms of universals, that is, the set of characteristics
that are
common to all objects that are members of this class (ibid.).
In the second -thou relationship. the thou is seen as an individual person. but "Gadamer
shows that this 'personal' relationship can still remain imprisoned in the I and actually is
a relation between the I and a reflexively constituted thou (as in "I feel your pain")
(ibid.). In this case. historical awareness "knows the otherness of the other. not in
relatedness to the universal" as was the case with the previous I-thou relationship.
but in
all its particularity. This is the kind of awareness that Gadamer critiqued, because it
claims to be objective and to fully know the other (ibid.).
The last type of I-thou relationship does not project the meaning from the . It "is
characterized by authentic openness to the thou" (ibid.). here, the 1. through his authentic
openness. "allows something to be said to him", to use Gadamer's terms. This "is the
kind of openness that wills to hear rather than to master" and. most importantly. "is
willing to be modified by the other" (ibid.). This kind of I-thou relationship is the one that
characterizes Gadamer's historical-effective consciousness.
This concludes the section on hermeneutics. With this background material, we can now
turn to the heart of the theoretical framework of this thesis. the taxonomy of
communication and integration practices in design and product development.
The Typology31
The objective of this chapter is to develop a typology of practices in design and product
development, using some of the concepts from linguistics and hermeneutics presented in
the foregoing sections. The practices we are concerned with are those communicative
practices whose objective is to provide tighter integration between the development
'3As discussed in Chapter II. this typology should be viewed as a cross between typology and taxonomy.
it is the result of a joint iterative process that involved both an analytical theoretical element (some of
which is captured in this chapter) and a significant body of grounded empirical work (some of it
described in the case study reports of the next three chapters.
90
as
organization and its customers. and between the different parts of the development
organization.
First, a preview of the typology is in order. As alluded to in the introduction, a key
distinction must be made between interpretive and non-interpretive practices. The latter
will be referred to as analytical or structural and. occasionally. mechanistic. Following
the example of linguistics, structural or non-interpretive practices could further be
divided into the syntactic and the semantic; at this point. such subdivision does not seem
useful, for reasons that are explained later. An important distinction will be made,
however, between two kinds of interpretive practices, those that will be called pragmatic.
and those that will be referred to as hermeneutic. This distinction is one of the key
contributions of this typology. What we have then is a trichotomy consisting of structural
("syntactico-semantic"). pragmatic. and hermeneutic practices.
It is worthwhile to contrast this typology with other similar ones, such as Charles Morris'
"famous trichotomy of syntax, semantics, and pragmatics." described in Akmajian et al.
(Akmajian et al.. 1995. p.343). That typology. discussed at length in the background
section on linguistics, sought to classify the different ways of studying sign systems in
general, and language in particular. (The same definitions that we looked at in the case of
language apply to other sign systems in general: pragmatics focuses on the way humans
or animals use the signs in real life: semantics studies the relations between the signs and
their meanings. and syntax studies the relations among the signs themselves, abstracting
from both meaning and use. (See Audi. 1995, p. 7 9 9 .)
Another similar typology is one given by Jantsch. where he applies the syntactic,
semantic, and pragmatic trichotomy to information (Jantsch. 1980. pp. 50-53). He
describes as syntactic the kind of information based on Shannon and Weaver's
information theory. which was touched upon in the introduction and briefly in this
chapter. in the context of the conduit metaphor of information transfer (Shannon &
Weaver, 1949). Syntactic information flows in one direction only (the functions of sender
and receiver are clearly defined). and the amount of information to be transmitted is
given
priori. and it can only decrease due to noisy transmission. Most importantly,
'new information is [...] primarily considered to reconfirm and strengthen existing
91
information structures" (ibid., p. 51). lie associates syntactic information with closed
systems. systems with fixed structures. Ilis explanation of semantic information is less
clear. Hie associates that type of information with self-organizing systems that are open
and therefore capable of getting rid of entropy. and he describes it as being exchanged in
a circular Cashion. and "within the context of a particular meaning" (ibid.). Finally he
describes pragmatic information as going beyond semantic information in the sense that it
is "geared to make a certain etffect." It is information that changes the receiver. lHe
associates this type of information with autopoictic systems, that is, seltf-renewing
systems whose structure is not necessarily tixed (ibid.). Whereas syntactic information
was characterized by a high degree of confirmation and little novelty. semantic
information combines both novelty and confirmation in "equal" amounts. U[nlike the typology to be presented here. all these trichotomies distinguish between the
syntactic and semantic levels or types. and relegate all else to the pragmatic level. At this
point. a word of caution is in order. concerning the proposed distinction between
pragmatic and hermeneutic levels. That distinction is not a natural one,. as pragmatics and
hermeneutics come from different traditions. as mentioned earlier. Yet. in his paper
comparing pragmatic and hermeneutic interpretation, Dascal ascribes to pragmatics the
same objective as hermeneutics (using (Gadamer's words): "Pragmatics. too. purports to
explain 'what always happens' in the process of interpretation. In some respects. its
account is similar to that of hermeneutics" (Dascal. 1979. p. 250). This makes
distinguishing between the two a difficult task, a fact underscored by the title of a recent
paper by Coyne, in which he refers to "Heidegger's pragmatics" instead of Heidegger's
hermeneutics (Coyne. 1998). Furthermore. it should be pointed out that one of )ascal's
objectives, in the above mentioned paper. in comparing pragmatics and hermeneutics is
"making each more assimilable by whoever belongs to the other tradition" (Dascal. 1979.
p. 240). Later on in the paper. having identified one aspect (among others) of
hermeneutic interpretation' that distinguishes it from pragmatic interpretation (at least as
conceived at that point in time). he goes on to explain that "nothing prevents pragmatic
2 Compare
to the pragmatic and hermeneutic notion of understanding presented earlier. where
understanding of the novel or alien could not take place w ithout the interpreter bringing his preknowledge (the old) to bear on the situation.
92
theory from appropriating this insight in its own terms" (ibid.. p. 248). It is therefore
important that the reader understand that the terms used in the typology presented herestructural, pragmatic, hermeneutic-are
being used in the specific technical sense set out
below. The linguist interested in evolving and expanding his field may wish to see these
distinguishing features co-opted into that field: for the purposes of this research. however.
it is important to maintain and underscore these differences.
Analyticalor structuralpractices
The first classification of design and PD practices is inspired by the positivist theory of
language that underlies the models oft'syntax and semantics presented earlier in this
chapter. These practices are either directly based on such reductionist models, or reflect
such a reductionist, Cartesian worldview. Because of their characteristics. and because
the term "practice" conveys a sense of situatedness that is not applicable to these
analytical approaches. it may be more accurate to talk of tools and methods: for the sake
of simplicity, however. I will continue using "practices"'. and sometimes "approaches". A
number of the important tfeaturesthat characterize these structural or analytical practices
are listed below.
As with semantics. structural approaches to PD are based on a particular model of the
system to be understood and analyzed. That system is seen as closed or self-contained. in
the same way that semantics contains its own fixed lexicon.
33
Any changes to the lexical
entries as the language evolves are external to semantics per se. At a minimum, the
system is fully defined and clearly circumscribed at any given point in time and for the
duration of the analysis.
Structural or analytical approaches follow the example of semantics, where the lexical
meaning of a sentence is derivable from the lexical meanings of its component words and
the rules for combining and ordering them. Similarly, in structural approaches to PD, a
problem or situation can be easily reduced to its constituent or atomic subproblems. The
manner in which these parts can be combined and ordered to form the whole and the way
'
Whether the lexical entry for each discrete unit (word) refers to its part-of-speech property or to its
lexical meaning is unimportant for our purposes. because we are abstracting from syntax and semantics
their high-level characteristics.
93
in which they interact follow clear, specific governing rules and constraints; in other
words, going from parts to whole is not problematic (see next paragraph.) Very often the
structural approach is presented as a methodology, in which case the relation between the
different steps and the progression from one to the next follow a specific ordering and
specific rules.
Since these are rules based on deductive or inductive logic, as opposed to, say. heuristics.
there is no room for ambiguity or contingency. The rules apply always, and always yield
the same repeatable outcome. No appeal to context or other external factors is needed to
reach an answer or a solution. We are dealing with universals. with objective knowledge:
the outcome is not dependent on time, on the particular individuals involved, their
history, their norms or their values, nor on the particular path followed. In other words.
the rules, their validity. and their applicability do not depend on the "content" of the
situation. In tact. it is by virtue of this repeatability (which the)y achieve at great cost) that
structural approaches can claim to have predictive power (at least within their limited.
circumscribed world), a feature to which they owe much of their attractiveness and
success.
Structural or analytical practices and approaches to PD have their basis in Cartesian
thinking, with its clear distinction between subject and object. These approaches are not
situated in human action. but they are conceived by the subjective mind acting against
clearly identified external objects. The stance of the thinking subject is therefore the
objectivity of the natural sciences. There is no place tfor prejudice (prejudgment) and preunderstanding, even if partial, and there is no place for reflection.
The model for these analytical practices is that of scientific theories which "state
deworlded relations between deworlded data", to use Heidegger's term (Dreyfus. 1991, p.
207).3 4 Unlike the study of human as self-interpreting beings. which "requires
interpretation within thefid hermeneutic circle of shared significance," analytical
approaches. even when they involve the study of human beings as objects "requires only
34
Dreyfus quotes Heidegger using the term "deworlded" in Histor oj the ('oncept of Time. Bloomington.
Indiana University Press. 1985.
94
the circularity of working within a theoretical projection" (Dreyfus, 1991, p. 203).3 That
is why one can make a distinction between the circularity of parts-to-whole and whole-toparts of lexical meaning or system dynamics for example, which only require deworlded
theoretical projections, on the one hand, and the true hermeneutic circle needed to
understand ambiguous, open-ended situations.
Examples of analytical processes in design
There are numerous approaches to design that are based on a "scientific" methodology,
and that could be described as analytical or structural in the sense of the term presented
above. These approaches which fall under the heading of "design science", use models of
design that exhibit "kinship with the logical-empirical approach and the hypotheticaldeductive system of theory construction I ... ]" (Rowe. 1987, p. 163).
Typical of these approaches are those that model the design process as one of search and
optimization, using generate-and-test algorithms. As an indication of how removed such
approaches are from human values, practices and areas of shared significance, they do
not allow room for the "style" of the individual designer, something most people take as a
self-evident characteristic of design as a creative activity. "Style" is viewed simply as a
byproduct of bounded rationality and the resulting inability to reach the optimal solution
to the design problem (Simon. 1981, p. 150).
Another example of an analytical approach is the use of shape grammars to automate
design (Stiny et al.. 1978; Mitchell, 1990). In these approaches. design primitives
(geometric shapes that play a role equivalent to that of words in language) are specified,
along with rules that determine how these primitives can be combined to form coherent
structures (the equivalent of the rules of grammar that allow words to be combined into
"meaningful" sentences.) These systems make it possible, for example, for a computer to
generate Palladian villas (or, more accurately. villas in the style of Palladio) (Stiny et al.,
1978; Mitchell, 1990, pp. 152 ff), or various pieces of consumer electronics in the Braun
3
The term "projection" is used by Heidegger to refer to the way man's everyday coping is organized by
that is. his final point, not in the sense of final goal, but in the sense of the
hisfor-the-sake-of-which,
long-term purpose of our being that gives meaning to our lives (e.g.. being a teacher or a parent.) (See
Dreyfus, 1991. p. 186 and pp. 92. 95.)
95
style (Wallace, 1991a). 3 6 In these approaches, as in semantics, the primitives are given;
how they came to stand tor what they stand for is external to the system or the approach
itself. For example. the rules for generating Palladian floor plans consist of rules for
generating grids, tfor I-shaped, '-shaped, and cross-shaped central rooms. for generating
porticos, and so on. These rules are totally removed from the 16'h century notions of
status, wealth, elegance, style, comfortable living, formal entertaining, and all the other
contextual and background factors that must have impacted the architect's work (by
virtue of his being situated in and engaged with a world of shared significance with his
community). as he interpreted these notions tor his clients. Similarly. reducing the Braun
style to certain corner treatments and radii and to certain shades of black does not account
for the social and political environments. the philosophical belicts. and technological
tactors that came together at one point in time to set the tone for the early Braun style
(through the work of Hans Gjugelot and his students at the lJim school of'design)
(Liindinger. 1990). and how that style evolved to its more recent Braun adaptations. More
importantly., these approaches cannot inform nor guide the evolution of a design language
or grammar over time. as the environmental and contextual tfactors change. For example.
such a system would not be capable of' taking the design of' Braun coffee makers from
their traditional cylindrical forms of the 1970s and 1980s to the newer conical shapes of
the 1990s. 37
Interpretivepractices:pragmaticand hermeneutic.
The other two types of practices are interpretive in nature; they are respectively inspired
by the ideas from pragmatics and hermeneutics presented earlier.
The key difference between analytical (structural) and interpretive practices or
approaches is in the underlying conception or model of the world in which they are used.
'6 Stiny and Mitchell were able to use their system to generate designs for villas that Palladio himself had
never actually designed. but which were nonetheless thought to be his by experts on the subject
(Mitchell, 1996). That is why "villas in the style of Palladio" is a more accurate terminology.
7
Wallace explains that an important contribution of this thesis is to show how a lot of industrial design
work is mechanistic in nature and can indeed be captured by such systems, and that industrial designers
should focus their energies on the much more difficult task of developing new shape grammars (Wallace,
1991a, p.30: 1991b). Such a system "uses conventions observed in existing products to design new
ones," "by borrowing ideas and visual language from other known and related artifacts." The system has
the advantageof "freeing the industrial designer to partake in truly creative design" (ibid., pp. 143-144).
96
The former apply within a particular theoretical model of the "world" (or system or
situation), a necessarily reductionist, fully circumscribed abstraction or simplification of
fhe real world. Interpretive practices, on the other hand, have as their range of relevance
the real world of everyday life. It is the world of situated human beings interacting with
one another against a background of shared social skills and know-how that cannot be
reduced to an explicit set of facts, assumptions. beliefs, and principles, no matter how
complex (Dreyfus, 1998. p. 285).
A certain amount of shared background is not the only contextual factor necessary for
successful interpretation and communication. Interpretive practices. be they pragmatic or
hermeneutic, are dependent on several levels of contextual factors, as discussed earlier.
These contextual factors, like the shared background, are all "external" to the
communication itself. For that reason, in interpretive practices, what is unsaid, the
implicit, can turn out to be just as important as what is said or given explicitly. This is not
the case with analytical practices.
As a result of their context-dependence,
interpretive approaches tend to be open-ended.
Unlike the analytical/structural approaches which are after the one correct, optimal
answer, interpretive practices admit that whatever answer is reached at any point in time
is tentative and fallible. subject to modification or even radical change as 'new'
contextual information is revealed. An interlocutor involved in interpretive practices
therefore plays an active, reflective role. At every point. he is asking whether the
interpretation arrived at so far is to be accepted, or whether an alternative should be
considered (and why.) His counterpart involved in analytical practices, on the other hand.
who can hardly be referred to as "interlocutor", is passively receiving information or
merely applying mechanical transformations to it.
Although the two types of interpretive processes share these general characteristics, they
differ in some important respects on which the typology is based. These are discussed
below.
97
The differences between pragmatic and hermeneutic practices
The two types of interpretive practices or activities differ on several dimensions. Some
are primordial, such as the degree to which each conforms to or rejects the subject/object
dichotomy. Others, though somewhat less fundamental, may be of greater practical
relevance. (These less fundamental ones invariably reflect the more basic ones; for
example, the constitutive role of the interpreter in hermeneutic interpretation is connected
to the rejection of the primacy of the subject, and of the subject/object dichotomy, and
thus of the preexistence of a separate, self-standing meaning "out there".) The latter
include the nature of the role of the interpreter or interlocutor, the role of the subject
matter itself. and the breadth of contextual factors.3 8
In both types of interpretive practices, the interpreter plays an active role. In pragmatic
interpretation, that role is akin to that of a detective: during the course of the interpretive
activity, the interpreter or interlocutor actively gathers contextual information that may be
useful and relevant to recovering or getting as close as possible to the speaker's
intentions. In hermeneutic interpretation, on the other hand. the interlocutor's active role
is more creative in nature. as he plays a constitutive role, via his prejudices. in the process
of coming to understanding. In this case, there is no such thing as an original speaker's
intention, a preexistent idea or meaning that the speaker is trying to communicate. The
contextual factors are not subordinated to the process of recovering the speaker's
intention but they themselves (as interpreter's prejudice) are part of the creative process
itself. These two situations can be considered the extreme points of a continuum which
would include those commonly encountered situations where the interpreter detects an
idea that is not fully formulated nor clearly articulated by the speaker, and participates in
developing it further through the exchange that takes place between them (Dascal, 1979.
p. 248).
Since hermeneutic interpretation is not focused on recovering any preexistent speaker's
meaning or intention, the original communicative object-the
statement-"hardly
initial question or
plays a distinctive or focal role" in the interpretation process (Dascal,
38 The presentation of these differences draws heavily on Dascal's discussion of the same topic (Dascal.
1989).
98
1979, p. 252). It is merely a starting point in the conversation that takes place between the
interlocutors and more generally in the interaction between the human agents. It is simply
an entry point into the hermeneutic spiral, described previously, of ever increasing
consciousness. This is different from pragmatic interpretation in which the interaction
between the human agents is seen as a significantly more "purposive and rule-governed
activity" (ibid.). Instead of some starting point that is likely to be overshadowed as the
interaction proceeds, and some general direction that is as likely to change as the
hermeneutic process unfolds, pragmatic processes try to stay focused on the original
intention of the interaction. In a pragmatic exchange, although any understanding remains
in principle tentative, there is present the notion of an ultimate understanding that is
derivable from the initial communicative object, that is, from the starting point of the
interaction. And even though that derivation may rely on heuristics, analogy, intuition
and other forms of abductive inference, pragmatic practices remains less open-ended than
its hermeneutic counterpart. In other words, pragmatic interpretation discriminates
carefully between the contributions made by the different factors involved in the
interaction to the final understanding: the interlocutors (e.g., speaker and interpreter, or
interviewer and interviewee), the initial communicative object (statement or idea or
question). and the various elements of context.
The last difference between pragmatic and hermeneutic practices (or at least the last one
to be discussed here) involves the range of contextual factors that enter into consideration
during interpretation. This may seem surprising since, as we saw, both allow for the most
general contextual factors to be considered: general extra-linguistic 'background'
knowledge in the case of pragmatics; the concept of the 'question' and the role of the
interpreter's prejudice in the case of hermeneutics. Since hermeneutic interpretation does
not accord any privileged position to what the utterer may initially have in mind, then
'the question' that motivates the initial communicative object. and in light of which that
object is to be understood, "may legitimately be found well beyond anything the utterer
might have in mind" as well (Dascal, 1979, p. 251). Similarly, as the process proceeds
through the spiral of coming to understanding, the interpreter "should inquire into his
own possibly unconscious motives for, say, being interested in the [communicative
object] at all" (ibid.). Therefore, hermeneutic practices do not discriminate between the
99
roles played by the various factors that enter into the process of coming to understanding,
be they subconscious motives, historico-cultural considerations, or the original intentions
of the people involved in the interaction. In a sense, this last difference is closely related
to the previous one. As the speaker's intention loses importance in the process of coming
to understanding, then it is only normal that contextual factors that are not directly related
to these immediate intentions end up gaining in relative importance.
In general, it could be said that "each pragmatic principle ha[s], for (philosophical)
hermeneutics, an additional metaphysical import..." (Dascal, 1979, p. 251). For example,
consider the reflective element that exists in both. In a pragmatic exchange, the "secondorder critical or reflective step" involves "asking whether there are reasons not to accept a
given 'transparent' interpretation," based on the contextual factors available (ibid.). In
hermeneutic exchanges, on the other hand, we saw that reflection moves to a higher level,
from searching to the question that motivates the initial object that starts the interaction,
to probing one's own motives for participating in the exchange.
Examples of interpretive practices in design
In recent years, design researchers have come to see the limitations of the analytical
models of the design process of the 1960s. as exemplified by Simon's rational problem
solving and optimization theory, which was briefly mentioned earlier (Simon, 1981). The
result has been a shift in the research focus, with renewed emphasis on the designers
themselves, and what it is that they do as they go about their daily work, as opposed to
some abstract methodology. Sch6n's description of the design process as one of
reflection-in-action is a prime example of that shift to an interpretive conception of that
process (Schon, 1983). The differences between Simon's and Sch6n's models are
summarized in the following table reproduced from Dorst and Dijkhuis (Dorst et al..
1996, p. 255):
100
Table 111.3:4comparison of an analytical and an interpretive model of design: Simon 's rational problem solving
model and Schn 's reflection-in-action model (Reproducedfrom Dorst and Dikhuis, 1996)
Item
'Simon'
'Schin'
Designer
= information processor
(in an objective reality)
= person constructing
his/her reality
Design problem
= ill defined, unstructured
= essentially unique
Design process
= a rational search process
= a reflective conversation
Design knowledge
= knowledge of design
= the artistry of design:
procedures and 'scientific'
laws
when to apply which
procedure/piece of
knowledge
= optimization theory, the
natural sciences
= art/the social sciences
Example/model
Schon's view of the designer as a "person constructing his/her reality" is reminiscent of
the hermeneutic notion of the interpreter's ever-expanding understanding and
consciousness. Similarly, his description of the design process as one in which the
designer is involved in a "reflective conversation" with the design situation is analogous
to hermeneutic interpretation, where the interpreter is engaged in a dialogue with the
subject matter (the text or his interlocutor's utterance), a dialogue that is reflective
because it invites the interpreter to inquire into his initial prejudices and to modify them
in the process of raising his 'historical effective consciousness". Furthermore, Sch6n's
view that every problem is "essentially unique" points to the fact that every situation is
highly context-dependent, a key element of interpretive thinking. It is very similar to
Gadamer's view that in language use. untranslatability is the rule (Dascal, 1979, p. 242).
Finally, in contrast to Simon's focus on knowing procedures and scientific laws,
something that can be done equally well by any individual or even machine, Sch6n
focuses on the artistry of design and the related know-how, which highlights again the
important role played by the prejudice and judgment of the designer and the
particularities of the situation.
In their paper, Dorst and Dijkhuis test the descriptive power of these two models of
design, by looking at how closely each of them reflects the actual activities that designers
101
indicate they are involved in in real-time. 3 9 The authors conclude that Simon's model "is
particularly apt in situations where the problem is fairly clear-cut," and where the
designer had well-established strategies and procedures he could follow in solving the
problem, as would be typical of the embodiment phase of design (Dorst and Dijkhuis,
1996, p. 269).4 0 Schbn's model on the other hand "works particularly well in the
conceptual stage of the design process, where the designer has no standard strategies to
follow and is proposing and trying out problem-solution
structures." The author stress, as
a critical aspect of Sch6n's model of design as reflection-in-action,
the fact that it does
not sever "the close link between the content and process components of design
decisions" (ibid.). This is in contrast to the analytical approaches to decision making in
which "deworlded" (to use Heidegger's term), value-free methodologies are assumed to
apply regardless of the specifics of the situation.
Structural, pragmatic, and hermeneutic practices: a summary
This section sums up the most important features of the typology presented in the
preceding few sections. First, Table 111.3presents the differences between structural
approaches and their interpretive counterparts, grosso modo. without distinguishing
between the pragmatic and the hermeneutic types. Next, Table 111.4focuses on the key
aspects that differentiate pragmatic from hermeneutic interpretive practices; where
relevant, the corresponding aspects of the analytical practices are also presented.
39The protocol analysis that they encoded was based on I 5-second intervals. In protocol analysis. the
researcher relies on a verbal account given by the subject. in this case the designer. of the cognitive
activities in which he or she is engaged at every time interval.
40
In one well-accepted model of the design process, embodiment design is the phase that follows the
conceptual design phase. In the embodiment phase, the designer gives shape or form to the concept
arrived at in the previous phase, thus the term embodiment (Pahl et al., 1984).
102
Table 1!. 4: Analytical v. interpretive approaches: a summary
Analytical ("Structural")
Interpretive
Abstracted from reality:
"Real-life", "actual use";
Theory- / Model-based (assumptions,
simplifications);
Description of what actually
happens.
"Deworlded."
Clearly
. ~conside
Closed, self-contained system;
Clea circumscribed
with fully
Clearly circumscribed with fully
characterized interactions at the
boundaries.
"Open" in the sense that any number
of factors might have to be
Almost no limit to how remote
relevant contextual factors may be.
Reductionist; partitioning of whole
into parts, as well as ordering and
combining of parts to form whole
follow strict rules.
Rules not sufficient for partitioning
of wholes into parts and for
recombining parts into wholes;
contextual factors are needed.
Circularity of working within
'theoretical projection';
Circularity of 'shared significance';
Reflectivity;
Objectivity of the natural sciences.
Prejudice and (preljudgment play a
critical role.
Outcomes are repeatable and
predictable;
Ambiguous; "untranslatability is the
rule;"
Governing rules based on deduction
Context dependent: contingent on
and induction.
multiple factors;
Abduction, heuristics.
"Content independent", non-
Each situation is (potentially) unique
normative: rules and methodologies
and therefore requires its own
apply regardless of the specifics of
treatment and approach.
the situation.
103
Table 111.
5 Analytical, pragmatic and hermeneutic practices: keyfeatures
Analytical ~~Analytical
~Interpretive
("Structural")
SyntacticoSemantic
Subject /
Object
Universals /
particulars
Pragmatic
Cartesian; full
separation;
Grey area:
Some
Some instrumentality:
instrumentality:
Thinking subject
communication is a
free of prejudice
purposive act;
and preconceptions Careful discrimination
acting against
between the role of the
preexistent, selfdifferent actors and the
standing objects
contribution of the
different factors involved.
Ditto for primacy of
subject;
No actor (interlocutor)
occupies any privileged
position.
Universality:
repeatability
"Untranslatability is the
rule";
Understanding result of
possibly unique interplay
between the situation and
the particular context
Role and
breadth of
context
Hermeneutic
None;
Model, process, or
Model, process, or
methodology is
circumscribed,
self-contained.
Generally very important;
Separation between subject
and object fully rejected;
Every situation is unique.
Crucial; prejudice
constitutive of meaning,
only for
Yet, it is not always
necessary to go looking for
a meaning bedisambiguation;
"transparent' one;
Dynamic: prejudice
Quasi-static:
changes
as
changes
as interpreter goes
Quasi-static: changes as
through the hermeneutical
the exchange develops, but through the he
given at any point in time; spiral, even if the
interaction has stopped.
Multiple levels (see
Multiple levels (see
background section on
pragmatics);
Subservient to
communicative intention of
speaker.
Role of
interlocutor /
interpreter
Receive
information;
"Process"
Disambiguate situation by
gathering relevant
contextual information.
information
mechanically,
using fixed,
deductive /
inductive rules.
104
Intentions as well as
"second-order
motivational, historicocultural, functional or
causal explanations equally
relevant.
Constitutive; participates
equally in developing an
understanding of the
situation.
Reflection
Not applicable;
Epistemological
Value of any bit of
information given
by model, not
subject to
judgment;
A critical, reflective step is
More ontological;
level;
involved in understanding
a situation: "Does this
understanding make sense,
given the context? Or is
there a reason to look for
an alternate explanation of
this situation?"
Information
processing is
mechanistic; rules,
Expanding one's
'historical-effective
consciousness' changes
one's prejudgments, and
therefore one's being;
The consciousness
involves itself into its own
reflection, through the
fusion of horizons.
as well as when
and how to apply
them, given.
Role of
prejudice
(prejudgment)
None: objective
scientific
(Cartesian) stance;
Assumptions and
background knowledge
cannot be made entirely
explicit.
All assumptions
explicit.
Role of
conversational
object
Primordial.
Focus of
interaction,
purposiveness
Fully goal
Positive, enabling;
Plays constitutive role in
coming to understanding,
as the original horizon of
the interpreter.
Primordial;
Starts conversation
End point derivable from
More open-ended
it. -
oriented.
Interaction purposive;
Interaction more open-
Focused on original issue
ended; has general
direction, but can go
represented by initial
conversational object;
anywhere;
Focus on identifying
communicative intentions
of speaker.
Characterized by openness
of 'true questioning',
desire to know;
Yet subject to constraints;
Interlocutors taken over by
the 'game'.
...
The typologyat work in PD
In closing this chapter, I will use the typology presented above to analyze a few popular
product development methodologies or practices and, in the process, to understand their
limitations and their range of applicability. This will be similar to the way in which
different design approaches-search/optimization,
shape grammars, reflection-in-
action-were analyzed and discussed earlier, using this taxonomy. The particular
105
methodologies discussed below are stage-gate systems and the design structure matrix
(DSM).
Stage gate systems
A stage-gate system is a workflow model of the product development process, from the
idea stage all the way to product launch and beyond (Cooper, 1990). A stage-gate system
divides the development process into a number of phases of activities, the "stages", each
of which is followed by a decision point, called a "gate". At each gate, a decision must be
made on whether to proceed to the next stage, to return to the current stage for more
work, or to abandon the idea altogether. That decision is referred to as a
"continue/abandon/recycle" or "Go/Kill/Hold/Recycle" decision ("hold" is another
possible gate outcome, where the project is put on hold, either to await a change in
available technologies or market conditions, for example.) Each organization, upon
implementing its stage-gate system. will define for itself the activities that are to be
performed during each phase, as well as the inputs, decision criteria, and outputs for each
gate (O'Connor, 1994). The inputs are the deliverables from the current stage of
activities, and the outputs of the gate are the decisions described earlier, as well as the
approval of a plan of action for the next stage. Figure 11.2, taken from Cooper (1990),
shows a typical stage-gate process that consists of five phases.
Underlying stage-gate systems is a conception of product development as a process
analogous to any other, which therefore can be managed using standard processmanagement methodologies (Cooper, op. cit.. p. 45). In particular, Cooper compares
product development to the manufacture of physical products, which can be divided into
a number of work stations (analogous to the stages), with quality control checkpoints in
between (analogous to the gates.)
106
tU\-
rlD
irJI..i
Ilzu:
FAA
)
S" ft . ,., '
Tf l-a
Nt.-
-l
cnl
nMi...
1Ik,,S4%IW,
Rrs
CA&.:
c1uutKXI
laairm
in*3>xs
lop-slur141icn
xM
Figure 111.2:Stage-gate system model (from Cooper, 1990)
The most important aspect of this analogy can be gleaned from the following statement
by Cooper:
The way to improve the quality of outputfrom the process, of course, is to
focus on theprocess itself-to remove variancesin theprocess (Coopers,
op. cit., p. 45).
The approach to product development embodied by stage-gate systems has many of the
hallmarks of a structural method, as explained below.
First, the reductionist nature of the model is already clear from the production system
analogy, with the division of the process into fixed sets of tasks at each stage, and rigid
"quality control" or inspection requirements at each of the gates. Secondly, the work-flow
model is presented by Cooper as a self-contained system, with clearly-defined inputs and
outputs, such as the initial idea that starts the process rolling ("The new product process
is initiated by a new product idea, which is submitted to Gate 1, Initial Screen" (Cooper,
1990, p. 52).) Thirdly, critical contextual factors are similarly not accounted for by the
system itself but are considered to be external to it, given from outside the system, just
like inputs. These include the decision criteria that are used at every gate, criteria which
are decided a priori. Consider for example Gate 1, the Initial Screen gate. Cooper
describes it as follows:
Gate I is a "gentle " screen, and amounts to subjecting the project to a
handfulof "must meet" and "shouldmeet" criteria. These criteriadeal
107
with strategic alignment, project feasibility, magnitude qf the opportunity,
differential advantage, synergy with thefirm s core business and
resources, and market attractiveness (Cooper, 1990, p. 52).
The above description begs all the important and difficult questions in conducting real-
life product development: What makes a market attractive enough? How does one assess
the magnitude of an opportunity at such an early stage, and how precise does that
assessment need to be? Will all the participants agree on the issues of strategic alignment
and synergies? Is it OK to deviate from these criteria under certain conditions? When and
by how much? Does it depend on the particular project? On the business climate? And so
on. Similar issues can be raised regarding the descriptions of the other gates and phases
of the system (Cooper, 1990, pp. 52-53).
The fact that there is no room for considering any contextual factors in the stage-gate
methodologies described above is not surprising given their aspiration to mimic
production systems. Although organizations are expected to customize the stage-gate
model to meet their particular needs, the overall direction is still towards the
implementation of a universal methodology that would be followed by all projects. with
the aim of increasing repeatability and predictability, and of "removing variances in the
process" (refer to earlier quotation.) Such a structural system leaves no room for
individual reflection and judgment. Even when an allowance is made for the fact that
different projects may have different characteristics or requirements, the goal remains one
of standardizing the process, through the development of different stage-gate models,
each appropriate for a particular, standardized type of development project. Quoting from
Cooper:
Not allprojectspass throughevery stage of the model. [...] Standard
definitionsofproject typesare developedand based on project scope and
investmentrequired.Appropriateroutes are determined.foreach type of
project. The routingfor any project at the idea stage is decided in the first
gate (Cooper, 1990, p. 53).
Again, the types and decision criteria are laid out ahead of time and the path to be
followed is determined and finalized at the beginning of the process. thus eliminating any
context dependent or situational changes within the process.
108
The push for standardization and repeatability leaves us with a system that is rather
generic and devoid of real-world considerations and meanings (we already saw that the
meaning of such things as "quality", "strategic alignment", "economic viability" and so
on are external.) The system can best be compared to the systems of syntax and semantics
described in the linguistics background section above.
Perhaps one could even think of the generic model as a syntactic model, and the models
that are organization-specific as semantic models. These are not systems that tell us what
to do in real use, what actually has to happen; it is akin to saying that a sentence consists
of a preposition, an object, a verb, and a subjective clause; that does not tell us how to
write literature.
The limitations of stage-gate systems, which are the result of their structural nature,
become clear when one confronts the number of non-germane elements that need to be
added to them in order to make them more useful in practice and better able to achieve
the optimistic results ascribed to them, as in the following claims:
Stage-gatesystemsform one solutionto what ails manyfirms' new
product programs.Facingincreasedpressure to reduce the cycle timeyet
improvetheir newproduc. 'it rate, ' corporationsare increasingly
looking to stage-gate models [...] (Cooper, 1990,1;. 4).
And:
One study (Booz,Allen & Hamilton1982)found that [...]firms that
adopteda.formal newproduct processdid better [...] (Cooper,1990,p.
47).
One of these unrelated supporting elements is the use of project teams to carry out the
development process (modeled using a stage-gate process), along with the attendant
organizational changes that necessitates:
The implementationof stage-gatesystemsrequirescertainorganizational
changeswithin somefirms. For example,a project team approachto
organizingnewproduct projects isfundamental to stage-gateapproaches.
No longercanprojects be handedfrom departmentto departmentwithin
the firm [...] (Cooper, 1990,p. 46).
109
It is not clear that the link between a project team approach and the use of a stage-gate
system is as fundamental as indicated in the above quotation. It is not so difficult to
imagine a company using project teams without necessarily following a stage-gate
system. Conversely, it is not impossible to have a firm organized into functional
departments follow a stage-gate process. As a matter of fact, a study that looked at how
several companies actually used stage-gate systems in practice found that one firm did
not form multifunctional teams until the third stage is reached, in order to control costs
(multifunctional teams being expensive to set up):
This means that stage I and 2 activities,typicallyinvolvingtechnicaland
market assessments and business case analysis, are done infunctional
departments and compiled by a program champion (O'Connor, 1994, p.
189).
Another important supporting element in stage-gate systems. one that might also require
an organizational change, is the involvement of senior managers as gatekeepers. In
addition to being "senior enough to have the authority to approve the resources needed by
the project," these managers should form a group that is multifunctional and
multidisciplinary (Cooper. 1990, pp. 46-47). Here again, the connection between the use
of a stage-gate system and the involvement of such senior managers in the process is not
necessarily a natural one. We know from the work of other researchers that having such a
strong senior manager with access to resources involved in the development processClark's heavyweight product manager for example-can by itself improve product
development performance (Clark et al.. 1991). Conversely, deploying a stage-gate system
does not automatically include such involvement, judging by the structural description of
the system. In his survey, O'Connor found that top management involvement remains
"difficult to gain and enormously difficult to sustain over time," notwithstanding the
adoption of stage-gate systems (O'Connor, 1994. p. 188). And in any case, getting the
necessary resources allocated to a project in real-life is a much more complex and
demanding process than the trivial stage-gate prescription of having a senior manager as
a gatekeeper. Other projects undertaken by the firm contemporaneously will p sumably
also benefit from having an equally senior manager playing a similar role. A more
evolved type of stage-gate system, a "third-generation" PD process model, attempts to
110
deal more pragmatically with this and other limitations of the "second-generation" stagegate model described in this section (Cooper, 1994). (More on this below.)
The same type of argument can be made about another element that is presented as "an
important feature of stage-gate systems" but that does not seem to be necessarily intrinsic
nor limited to it, namely, "parallel processing" or the execution of activities in parallel
instead of sequential fashion (Cooper, 1990, p. 49). Ditto for other "benefits" of stagegate systems such as improved quality of execution of important PD activities and
increased market orientation (Cooper, 1990, pp. 47-49).
The limitations described above are typical of several approaches to product development
that attempt to reduce what is in reality an interpretive situation to an analytical process.
Another such approach is Quality Function Deployment, which was briefly discussed in
the Introduction (see Chapter I.) As O'Connor's survey shows, the successful
implementation of a stage-gate process is really a process of organizational change, with
all the difficulties and pitfalls that such a process, grounded as it is in human interaction
and organizational culture, entails. It is not merely a matter of having product
development projects follow a particular analytical process model. Rather, it involves
"motivating" people, "secur[ing] senior management commitment [...] and
involvement," "guid[ing] and influenc[ing] the organization," and so on (O'Connor,
1994, p. 197). The fact that the implementation of any PD process is intimately linked to
the particular context of the organization involved is captured in one of his conclusions:
Yet no one suggests, nor does any empirical evidence exist, that supports
the notion that one specific NPD process is better in all organizational
settings than any other process (O 'Connor, 1994, p. 199).
The "third-generation" NPD processes, briefly touched upon earlier, are intended more
closely to reflect actual, real-life product development projects. Specifically, they seek to
address the limitations that come from forcing all development projects into a "one-sizefits-all" model, such as the "second-generation"
process model on which we have
focused so far. These newer models allow for the fact that every project is different and
may require a unique treatment, a treatment that depends on the particular situation and
context. "Third-generation" process models allow for such context-dependency through a
III
number of features including "fluidity", "fuzzy gates", and "flexibil[ity]" (Cooper, 1994,
p. 9). "Fluidity" is meant to reflect the need for the process to be "adaptable": stages can
be overlapped, activities are no longer "married to specific stages", and so on. "Fuzzy
gates" feature "conditional Go decisions (rather than absolute ones), which are dependent
on the situation" (ibid.). The process is "flexible" in that "it is not a rigid stage-and-gate
system: each project is unique and has its own routing through the process" (ibid.).
The "third-generation" process model described above is fundamentally different from
the rigid, mechanistic "second generation" model covered earlier. "Fluidity" and
"flexibility" can be viewed as making room for the numerous contextual factors that enter
into the assessment of the quality of what has been accomplished at the end of every
stage. They also make it possible to account for the conversational aspect of the decision
making process at the every gate. Cooper closes his description of the major features of
the new process model by cautioning against a "possible negative consequence," namely,
"fallibility":
This Third-GenerationProcessintroducesmuchmore.freedomand
discretionto project leaders,teams, and senior managers(whoare the
gatekeepersor decisionmakers.) Withfreedomand discretion,of course,
comes risk: the odds of making mistakes go up (Cooper, 1994, p. 12).
The introduction of this notion of fallibility is the clearest indication that, in the end, the
PD process is an interpretive process that cannot be captured fully using structural
models. In the next section, we look at another methodology that has been developed for
the purpose of organizing product development projects, the design structure matrix.
DSM (Design Structure Matrix)
This section considers another tool used to organize product development activities, the
design structure matrix or DSM. The DSM methodology was initially developed by
Steward in 1981 (Steward, 1981). It was later picked up and further developed by
researchers at MIT, and has been gaining increased acceptance in industry at companies
such as Ford and Boeing, where it has been used in understanding and modifying the
structure of complex product development projects (Eppinger, Whitney, Smith, &
Gebala, 1990;. As with the stage-gate process, which was the subject of the previous
112
section, this section attempts to understand the range of applicability and limitations of
the DSM methodology using the framework and typology developed earlier in this
chapter. In the process, we will look at its evolution as practitioners have started using it
in increasingly realistic and more complex situations.
The Design Structure Matrix method was originally developed by Donald Steward as a
means of addressing a basic limitation of standard design/project management tools, such
as the Critical Path Method (CPM) and Program Evaluation and Review Technique
(PERT). Specifically, these tools are limited to representing independent tasks or
activities that can be performed in parallel, or dependent tasks that must be performed
serially (Steward, 1981, p. 71). They cannot represent interdependent activities, which are
related to one another in a circular fashion (what Steward refers to as "circuits".) Yet,
such activities are very common in design and product development, due to the iterative
nature of many design tasks and sub-tasks, and to the learning that takes place in design
through trial and error. Eppinger (1994) provides a very good summation of the concepts
and assumptions that underlie the DSM approach. To paraphrase:
a) The basic elements of the design process are viewed as tasks of some duration,
which require information as input and produce information as output.4 '
b) The design process can be improved (in terms of speed and efficiency) by resequencing these tasks such that available information is put to use as early in the
process as possible, and information is made available to tasks that require it as
soon as soon as possible (Eppinger, 1994, p. 11).
In Steward's original work, the focus was on the variables or parameters that define the
design, manufacture and behavior of the product, and with the relationship between them
(Steward, 1981). In the more recent work, the definition of the unit of analysis in a DSM
has been broadened to include design tasks (Eppinger, 1991; Eppinger et al., 1990;
Epppinger et al., 1994). product development teams (McCord & Eppinger, 1993), as well
as subsystems and system components (Pimmler & Eppinger, 1994).
41 As
we will see later in this section, DSM can be used with different units of analysis than tasks.
113
From an operational perspective, the DSM methodology consists of the following steps
(using a task-based DSM as example, without loss of generality.) The first step is to list
the component elements or tasks, and to identify the precedence relationships between
them. Depending on the information flow between any two tasks, they can be classified
as dependent, independent, or interdependent. Figure 111.3below, taken from Eppinger et
al. (1994), illustrates these relationships using directed graphs to indicate the
informational flow between two tasks A and B.
Dependent Tasks
(Scics)
IndependentTasks
(Parallel)
Interdpedent
Tasks
(Coupled)
Figure 11.3. The relationships between design tasks (from Eppinger et al., 1994)
In the next step, these tasks are arranged along the identically labeled rows and columns
of a square matrix, and entries in the body of the matrix are used to indicate the
dependence of the row elements on information from the column elements. Figure III.4
(also from Eppinger et al., 1994) shows such a matrix for a project consisting of twelve
tasks. The third step involves re-sequencing these tasks, that is, interchanging rows (and
their corresponding columns) with the goal of making the matrix lower triangular. In that
ideal but rare situation, no task would depend on information generated by a downstream
one, and no iteration would be required. The more typical outcome is a block triangular
matrix, where interdependent tasks are concentrated in a few, relatively small blocks on
the diagonal. Figure III.5 shows one such re-sequenced task matrix for the same project
as in Figure 111.4.
114
A
C D E F G H
C
J K L
X
A X
D ..-------- .)..---------------E
F
G
HX
X
X
·X
X
X
*
X
X
XX
K
X
X
X
X
X
X
X
XX
X X
LX
XXX ·
Figure 111.4:The original design structure matrix (from Eppinger et al., 1994)
Figure 111.5:The DSM after re-sequencing of tasks and partitioning (ibid.)
Interdependencies are now limited to the smaller blocks on the diagonal. Within these
blocks, it is relatively easier to start the iteration process; furthermore, certain "tearing"
procedures have been proposed for identifying those parameters or task output
information, the removal of which (through the use of reasonable initial estimates for
example) allows the rest of the block to be made lower triangular (Steward, 1981).
Eppinger et al. (1994) extend the DSM methodology through their introduction of the
numerical DSM. The entries in the matrix are no longer limited to marks indicating strict
precedence relationships; instead, numerical values are used to represent the strength of
the dependency between tasks, or any number of other measures. Along with the
115
expanded representation, more sophisticated analytical algorithms are used that take into
account that information when reorganizing the tasks.
Early research: DSM as a structural tool
The rationale behind their DSM work, according to Eppinger et al. (1994), is the
argument that the design methodology used in any given design situation should be based
on the "underlying structure of the design problem." The authors trace this argument not
only to Steward, but also to Simon (1970) and, before him, to Christopher Alexander's
form synthesis methodology (Alexander, 1964). The authors believe that the structure of
the design procedure should depend only on the technical nature of the design problem
itself:
The analysiswe will performconsidersthe relative importanceof each
d signparameter to otherparameters,allowing the information
requirementsto determinethe appropriateschedulingof the decisions
(Eppinger et al., 1990, p. 40).
The method proposed by the authors for designing a design procedure is an excellent
example of an analytical or structural process, as described in Table 111.3above. First, the
proposed method is "deworlded", in the sense that it is based on a "mathematical or
engineering model of the system to be designed" (ibid.), a model that is abstracted from
reality. More importantly, even though the authors include the process of developing the
model as one of the steps in their methodology, they make no mention of the assumptions
and simplifications inherent to that process, and the attendant judgment calls. In fact, they
imply that there is one model for a given system ("Make the mathematical or engineering
model of the system to be designed" (ibid., emphasis added.)) This is an obvious
reflection of the circularity inherent in working within a theoretical projection.
The method is clearly reductive in that it relies on breaking up the design into its
constituent tasks, and the relationships between tasks into one of three clear, crisp
categories of information transfer. The method not only operates within a closed system
with clearly circumscribed boundaries, given by the model. It is also content independent:
the content of the labels attached to the rows and columns of a DSM are unimportant;
whether a task is labeled "A" as in the figures above, or whether it is labeled "Check
116
Geometry Markup" (Figure 5 in Eppinger et al., op. cit.) is immaterial to the procedure
itself.
The method as proposed by the authors is effectively devoid of human judgment or
prejudice. For example, when designers are interviewed in the early phase of the process
of developing the DSM, the objective is to obtain the relationship between design
parameters:
To create a parameter-leveldescription,we documentthe designprocess
by interviewingengineersonly (not their managers).We ask the designers
whichparametersmust be knownin order to set anotherdesignparameter
(Eppinger et al., 1994, p. 7).
Later in the process, after a new design procedure is developed, its goodness is assessed
analytically, not by human designers or managers. A mathematical sensitivity analysis is
proposed for assessing the impact of errors or changes in specifications on the progress of
the design process, and simulations of the design sequence are used to time it (Eppinger
et al., 1990, pp. 42-43).
The purely analytical orientation of this early DSM work is well articulated in the
following quote from Eppinger et al., (op. cit., p. 40), which makes it clear that there
should be no role for historical contingency nor for the particular organizational context
in determining the proper design procedure for a given product:
The benefit of this workfor the second type [of design. the redesign of
existing items,such as automobiles,]may bejust as great but harderto
recognize. This is due to the fact that a "procedure " exists and seems to
work well. However, it may have grown up [sic] organically and
historically and may never have been subjected to careful analysis. So its
internalinefficienciesor irrationalitiesremain undetected.
Having established the analytical structural nature of the early DSM research, I turn next
to more recent work that takes a more interpretive approach.
The interpretivenature of more recent DSM research
In some of the more recent work on DSM, the focus has shifted from the mechanics of
the methodology and from problems where the technical structure of the artifact alone
117
determines the DSM, to the process of properly reflecting the knowledge, experience, and
judgment of engineering managers and experts. An example of this more recent DSM
research is the work of Dong, which is summarized in one section in an MIT DSM
tutorial (Dong, 1999; Eppinger, Whitney, & Yassine, 1999). The section in question is
entitled "A Proposed Approach for building credible DSMs."
The first step in the proposed approach is to "Define the System and its Scope." This
critical step, which was taken for granted in the earlier work, is an indication that these
researchers are more aware of the fact that they are working with a model of the real
world, a model that is clearly circumscribed by its boundaries. As they point our, the
output of the DSM analysis depends on how the system is defined and on where its
boundaries are drawn (Eppinger, Whitney et al., op. cit.).
Step 3 in the proposed approach involves studying the information flow between the
system elements, which were identified in an earlier step. What is notable here is the
recognition of the importance of conducting face-to-face interviews with the engineers
who hold that knowledge, in addition to reading design documents. This is to be
contrasted with the earlier work, where the focus was on deriving these relationships
from engineering models of the artifact being designed. More importantly, the author
points out that different engineers will often hold different views concerning the
relationships between the different elements in the model and their importance. One
reason mentioned for these differences between engineers is "the different perspectives
on the issues due to the difference of their work." This is recognition of the role prejudice
or prejudgment plays in people's understanding (see Table 111.3above.)
In Step 5 of the proposed approach, one of the goals of DSM is described as "to aid the
design engineers and engineering managers to understand the design process better" and
to communicate better. The DSM is presented as playing the role of a boundary object,
the purpose of which is to help different people with different prejudices and worldviews
come to a shared understanding, giving the DSM a distinctly interpretive role (Carlile,
1997).
In closing this section, it is interesting to note that Steward clearly understood the
limitations of the tool he was proposing, the limitations uncovered and discussed in this
118
section. In his seminal paper, he notes, early on, that the precedence relationships
between design parameters "do not have to be based on mathematical equations", that
"[t]hey could represent qualitative judgments by managers or engineers", based, for
example, on their perception of the risk associated with a particular solution (Steward,
1981, p. 71, emphasis added). Even more interesting, and of particular relevance to this
discussion, is the linguistic analogy Steward uses to describe the limitations of the DSM
as a tool:
The precedence matrix represents the structure of the system, i.e., what
affects what. Semantics is the "how" and "why " of these effects.
Choosinga good tearing involvesan interplaybetweenan analysisof this
structure, which can be done with the help of a computer, and engineering
judgment about the semantics, which is done by a person (steward, 1981,
p. 73: emphasis in the original).
(The selection of a "good tearing", as mentioned earlier, involves choosing, from among
the unknown but needed design parameters, those for which reasonable estimates are
easily obtained.) Steward's use of the term "semantics" does not refer to the structural
semantics discussed earlier in this chapter, that is, to a lexical set of meanings and their
relationships. Rather, his "semantics" is closer to our pragmatic interpretation, as he is
referring to an activity that draws on the judgment and intuition of the engineer, and on
his experience with similar design problems or perhaps with different problems within a
similar context.
Finally, it is worth noting that in his "Notes on the Synthesis of Form", which is often
described as the seminal work in the area of structural design methodologies, (viz. earlier
reference), Alexander was actually concerned with finding a good fit between the
designed artifact and its context (Alexander, op. cit.). This concern and his approach
concern appear closer to our description of a pragmatic interpretive process than a
structural one, especially in view of his expansive notion of context-"anything in the
world that makes demands of the form" (ibid., p. 19). His proposed approach revolved
around identifying all possible sources of "misfits" between the form and its context, be
they functional or economic requirements or constraints, and, in a second step, to cluster
these variables in subsets based on the density of the interaction between them. His
119
description of how these interactions might be identified, which is quoted below, leaves
no doubt as to the interpretive nature of his approach.
The searchfor causalrelationsof this sort cannotbe mechanically
experimentalor statistical;it requiresinterpretation:to practice it, we
must adopt the same kind of common sense that we have to make use of all
the time in the inductive part of science. [...]
We shall say that two variables interact if and only if the designer can
find some reason (or conceptual model) which makes sense to him and
tells him why they should do so (ibid., p. 109). (Emphasis in original.)
The notion that a better solution to a design problem could be derived solely from a
mathematical model and a set of sensitivity analyses seems at odds with Alexander's
thinking.4 2
42
In the preface to the paperback edition of his book, Alexander makes clear his concern about design
methods that remove or abstract the process from practice (Alexander, 1964).
120
Chapter IV: Case Study I
Chrysler Corporation
The Chrysler Corporation was selected as a site for a case study because it has been at the
forefront in adopting several organizational innovations in the area of product
development. Examples are the use of small, dedicated platform teams, and the early
involvement of, and increased reliance on, outside suppliers. Both of these organizational
approaches are featured prominently in this case study. The study centers on the
development of the current Chrysler minivan, which falls in a product category that
Chrysler pioneered and in which it has enjoyed unmatched market success starting over a
decade and a half ago.4 3
Another important factor in selecting Chrysler as a research site is the company's leading
role in design, what used to be referred to in the industry as "styling". Chrysler has either
pioneered or capitalized on a number of recent trends in the industry, such as "cabforward" design and retro or "heritage" design. The company has received critical
acclaim and a number of awards from the Industrial Designers Society of America for its
designs, one in particular in 1995 for the minivan that is the subject of this case
(Woodruff, 1995).
The NS Minivan Team
The Chrysler team tasked with developing the Minivan replacement faced an
overwhelming challenge: how to come up with a product that would allow the company
43 "Pioneering" here refers to the modem minivan, which is typically based on a front-wheel drive platform.
There were other minivans that preceded the Chrysler design of the mid 1980s, such as the Volkswagen
Bus.
121
to maintain its dominant position in the minivan segment, at a time when more and more
competitors were entering that market. Chrysler had practically "invented" the modern
minivan and, over the years, the minivan had become its most visible mainstream product
and one of its most profitable. In the words of one informant, that product was the "heart
of the company"; it was "holding the company up..." Achieving continued leadership not
only required that the development team correctly identify and preserve those aspects of
the original design that made it such a success, but also that they find new features and
new minivan characteristics that would make the new model a runaway success. The
team also had to be inventive enough to anticipate future market pressures, from a
number of motivated competitors. One informant compared the situation to being "the
first sailboat in the sailboat race, which means you're out in front and if you screw up,
everybody goes in the right direction except you."
Beyond the uncertainty about customers' needs and desires, and the uncertainty about
competitors' future moves, a number of factors contributed to making the development of
the next minivan more arduous. One such factor was Chrysler's intention to increase
minivan production volumes beyond what the company had been accustomed to. Another
factor was one of the main sources of Chrysler's success at that time, namely, its
innovative styling. The company had recently introduced a number of models that were
acclaimed for their styling; styling that was seen as aggressive and daring in the case of
the Viper sports car, and clean and elegant in the case of a two-door convertible. Even the
functional LH sedans had received significant favorable press coverage for their "cabforward" styling. Yet, given its utilitaiian nature, the minivan is the one vehicle where
exterior styling plays a secondary role, and where interior packaging is much more
important. "This is the vehicle, more than any other, that gets designed from the inside
out." GM's Pininfarina-designed minivans, the so-called "dustbusters". were mentioned
as an example of a failed design where too much attention was paid to exterior styling, at
the expense of interior space utilization.4 4
In the face of these difficulties, the minivan team responded in two seemingly
contradictory, or at least inconsistent, ways. The first response was for the team members
44
The Pontiac Transport. introduced in 1989.
122
to become more customer oriented, "really starting to pay attention to the voice of the
customers, so we really became tremendous customer advocates." This focus on the
customer was reflected in the number of market clinics that were held around the country,
and in the number of intricate and detailed QFDs or "Houses of Quality" that were
developed during the course of the project. At the same time, the team's second response
was to rely on what one informant referred to as "the nontraditional kind of decision
making", what at various points team members described as following their "gut feel". A
significant part of this interview was spent trying to unravel this seeming contradiction,
and to understand the interplay between these two responses.
His master's voice: VoC,QFD, and all that
This intense focus by the minivan team members on the voice of the customer was not
meant to imply that the company had not been pro-customer during previous
development projects. Still, it is clear from the informants' comments that customer input
was treated differently on this project. Some of these differences were quantitative,
relating to timing for example. A thorough customer study was conducted during the very
early phases of the project-it
was described as "being in an extremely proactive mode
with the minivan", and "really doing our homework up front"-and this study was of
"extreme importance" as the project progressed. Another example is the number of
consumer clinics that were held and the number of QFD's that were used.
Other differences were qualitative in nature. For example, to some of the informants, this
level of customer involvement in the design process was clearly a novel experience:
We did create a lot of questions that were used in the clinics to try and
help clarifyisome of the customer requirements, which was probably, quite
fiankly, at that point in time, a little bit new. I know we weren 't used to
doingthat exactlyas engineers.[...] We 'reused to asking questions,but
we weren 't used to asking those kinds of questions.
Perhaps the most significant qualitative difference was the seriousness accorded to the
voice of the customer during this project. Statements such as the following indicate that
some of our informants might have been somewhat leery of customer input previously:
123
And, the topfive of those [issues],ifyou sat aroundthe table with any
given cross group, ifyou went around those,you basicallygot head nods
around the table. It becomes self-evident that this is... obviously what
they're telling us is true. We can all relate to this. It makes sense.
Chrysler had a significant advantage compared to its competitors when it came to
collecting customer input. As a result of its leadership position in the minivan market, it
had a large pool of customers to draw upon. For example, the team had a number of
unsolicited letters from minivan customers, which were carefully scrutinized early on in
the development process. The team also had available to it a relatively large number of
press reports.
A particularly important source of information, one that was mentioned during different
interviews, was the large number of minivan users within Chrysler itself. In particular, a
large percentage of the workers at the minivan plants drove what they built. These
customers, who were very motivated to see the next minivan succeed, were surveyed and
interviewed in depth on several occasions. More importantly, they were not only asked to
answer questions on use patterns or to comment on proposals; rather, the development
team actively sought their ideas and suggestions for improving the product, effectively
bringing them into the development and innovation process.
"Gut feel" and glory
Amid all this discussion on the importance of market research and of listening to the
customer, the informants would bring up, on several occasion, the importance of
following their "gut feel" in certain decisions. The archetypal example centered around
the decision to offer a fourth door as an option on the minivan. That is, a second sliding
door on the left side of the minivan, something that previous models, whether from
Chrysler or from its competitors, had never had. One informant, expanding on the
previously mentioned "nontraditional kind of decision making":
Thefourth door was kind of classicin that regard,I mean,you didn't
reallyhave reallystrong... You had a gutfeel that was differentthan
maybe what the market research was telling you, you know, in terms of
what you would sell. You ended up making the decision on your gut feel
more than on traditional market analysis.
124
It was not only formal market analysis that argued against the fourth door. Another
informant described his personal experience with friends and neighbors, who reacted
negatively when they saw the fourth door on a disguised prototype he was driving.
... people go, Oh, my God, you can't do that! There's a reaction that says,
you don't want that,you don't wantyour kidsjumping out. That'sthe
traffic side! Don't go there! This will never sell kind-of-thing.
It turned out the decision to go with the fourth door was a good one. Not only were 85%
of Chrysler new minivans ordered with that option, but competitors had to scramble to
respond, first by offering rebates on their otherwise excellent products (the Ford
Windstar, a "credible product", was made "obsolete", in the words of one of our
informants), then by redesigning their products to retrofit them, at great expense, with a
fourth door option.
Innovation and the conflict between VoCand "gut feel"
The conflict between listening to the voice of the customer and listening to one's "gut
feel" showed up mainly in situations that involved innovations, ideas with which the
development team did not have previous experience, and for which market data was not
available. Consequently, the conflict was central to the struggle to differentiate the new
minivan from its competition. The development team members were very aware that they
could not hope to come up with the necessary innovations simply by relying on customer
input and data. This point was made perfectly clear by one informant as he explained how
decisions driven by QFD and those inspired by "gut feel" played different roles in the
development process, and how one complemented the other.
One [QFD]is based on datafrom the customerin a unique methodical
translation of what we feel the customer wants [...]. The other one ["gutfeel"] is... Ifyou were to do that, everybody would have the same product,
because everybody listens to the same voice of the customer. [...] But how
you executethat and howfar you carry the design, and how innovative
you want to be, how anticipative you can be, that's what the difference is;
and how [much of a] risk-takeryou are.
(The connection between innovation and risk-taking is discussed in a later section,)
125
It should be pointed out that this notion that design by market research is incompatible
with innovative design is an old one. Critics and the specialized press will often refer to
an unimaginative or unexciting product as having been designed by marketers who lack
passion for the product. By contrast, our informants mentioned passion several times
during this interview, which is remarkable considering that these were mostly engineers,
not designers, and the project in question involved a minivan. not some high performance
sports car. How were the minivan team members able to combine market research and
the voice of the customer on one hand, and their "gut-feel" and self-professed passion for
the automobile on the other? The decision concerning the fourth door is instructive here.
The answer to the dilemma: interpretation
It would be unfair to say that using customer data and relying on "gut feel" were two
mutually exclusive decision making modes. each one applied under different conditions.
In fact, in all decisions relating to the minivan project, both modes appear to have been
involved to varying degrees. This makes sense: it would be difficult for someone to
simply turn off one's intuition or emotions when looking at a set of research results.
Furthermore, on most issues it appears that "gut feel" and the customer data were in
accord; these were the easier decisions, as explained by one informant:
And, if the gut supports the QFD, then you're pretty comfortable. When
the gut and the QFD are a little ways apart, you don't know which one to
believe, you can either believe your gut, and throw a whole bunch of
money and take a chance, or you can believe the QFD and be sqfe. When
you got that conflict, you've got to kind of say, something's going on here,
we've got to understand this, and then you've got to dive in a little more...
It is the cases of conflict that are more instructive and that are the focus of this section.
What did it mean, in the case of the fourth door, "to dive in a little more"? This question
was answered by another informant:
... that kind of goes back to gut, which you use. You can take it back to
these top ten items, what is the vehicle for? And what you find is, utility;
and knowing the vehicle and knowing the customer, knowing what you do
with it every day, and having a team that's used to using these things,
saying, now, f you had a door there andyou couldjust put your stuff in
every day, and you can think often examples in the last week where you
126
could use that, it all fits back with the QFD, it all fits back with your
experience every day, and everybody sees that there 's almost no negative.
You put a child latch on it, it takes care of the people who don't want
people to be able tojump out, andyou've got almostat somepoint [to]
say, Hhow could this be wrong? Andyou've got to go.forward...
In order to resolve the conflict, the team members went back to their own understanding
of the basic purpose of the minivan as a vehicle; what they latched onto was utility.
Instead of trying to resolve the conflict at the level of the feature itself (whether to have a
fourth door or not, whether to use a sliding door or hinged one), they moved the
discussion to higher level of abstraction, to the realm of the customer's needs and wants.
They then interpreted those needs and wants in their own way, which resulted in the
designers including a sliding fourth door as an option.
In doing so, the team members were substituting their own preferences, their notion of
what the product should be like, to those of the customer. Effectively, they were saying to
the customer, "You don't really know what features you need on this vehicle; just tell us
what you want to do with it, how you want to use it; we'll tell you what you need." Such
an approach does not accord with the prevailing analytical view of product development,
where the customer is assumed to know best, and where the designer's job is to generate
a verbatim translation of the customer's wishes.
Such interpretation can only be successful if the interpreter, unlike a mechanical
translator, has a thorough understanding of the worldview of the speaker whose
utterances he or she is interpreting. In other words, the designer or design team have to
thoroughly understand the background, concerns, and interests of the user of the product.
From the above quotation, it appears that the Chrysler team members did indeed achieve
such a high level of oneness with the customer ("knowing the customer") and, in many
cases, effectively lived the life of the typical minivan user.
It is only in this context of sharing the life of the customer that certain of the informants'
statements begin to make sense. For example, the following comment about the
emotional side of understanding the product and listening to the customer:
I think we had, as a group, we had a real strong sense of theproduct,
that's hard to put on paper. Itjust doesn'tcome out--andthat's wherethe
127
gut comes in--itjust doesn'tcome out in the QFDand all the rest of that
stuff Those are, they're good processes, but they 're a little bit mechanical
too, in their own way, and there's this emotion side of it that when you
listen to what people are saying...
External integration between the designer and the customer is only one aspect of
successful product development. The above statement also points out the importance of
the fact that the whole group had a "shared sense of the product". In the case of a
multidisciplinary development team, this internal integration is another key ingredient for
successful interpretation. The findings at Chrysler regarding this issue are presented next.
What makes interpretation work in this case? The team
By definition, acts of interpretation draw on the interpreters' worldview, background
knowledge and understanding, and his or her assumptions about how things function and
what lies at the root of empirical observations. Interpretation is also intrinsically
connected to the interpreter's concerns, interests, and immediate goals. They are therefore
highly personal and subjective acts, and one person's interpretation of a given situation
will generally differ from another's, to varying degrees.
This subjective aspect of interpretation can be the cause of disagreement and friction
when a number of people are brought together to work on a situation involving
interpretation; in particular if these individuals have different technical and professional
backgrounds and goals (Clausing's "dysfunctional specialization" (Clausing, 1993).)
Chrysler's minivan team members appear to have avoided such problems due to the high
degree of cohesiveness within the team.
There are two at least two conceptions of how a team functions. In the first, the team is
seen as a set of individuals, each with a different set of beliefs and a different perspective
on the situation at hand, who come together to perform a given set of tasks. Even though
these individuals all have the same goal, they never really develop a common vision of
what the solution ought to look like. The team then becomes a forum for them to
compromise, each to a varying degree accepting less than what he or she believes the
ideal solution should be.
128
The other conception of the team is one in which the individuals end up developing a
shared perspective on the situation and a common vision of the solution. The team then
becomes an integrated entity with many of the characteristics one would ascribe to an
individual, such as a distinct personality, convictions and passion. In that case, the team
members are no longer searching for a compromise solution, the one having to do with
less in order that the others can also get some of what they want. Rather, each team
member becomes equally concerned with making sure that the needs of the others are
met, because each team member understands the importance of the others' contributions
to the overall solution, which is also his goal.
When the minivan project first started, Chrysler's reorganization into dedicated platform
teams was close to being complete, and by then the company "had worked some of the
bugs out" from that organizational model. According to one informant:
... it was the.first time really that all the disciplinescame togetheras a
team [...] so when decisions were made. everybody bought ito it. There
was a great deal of compromise,but there was no arguinglater...
Although the term compromise was mentioned here and in other statements (e.g.,
"Sometimes, between two competing parties, you have to strike a marriage of
convenience..."), other statements make it clear that a higher level of integration was
present within the team. In particular, the engineers on the team were not only thinking in
terms of the technical aspect of their solutions; but they were equally concerned about the
marketability and the financial viability of their solutions, as well as the long term
success of the company. The following statement by the Executive Engineer on the team
is typical:
Everything we do is... we're in business to delight customers and make
money. I think, we're not in the business to do one or the other. we're in
the business to do both because I think you've got to do both to kind of be
successful over the long haul. [...] we want to be a nice profitable well-
thoughtof company-for a long time.
Another statement, concerning his interaction with the finance staff, is also indicative:
... the few times I had dealings with them [the finance people], it was them
tryingtofigure out how to get me to do somethingthat,you know, I didn't
129
really want to do necessarily. But as we kind qf'reorganized our thinking
as a company in the dedicated platform teams, the thing that became
instantaneouslyclearwhen they did that was what an allyfinance really
was, or is. They were the most cooperative, the most knowledgeable, the
most helpful group of'people that I ever worked with. [...] and it' not like
they weren't that way before, but it'sjust kind of the alignment qofthe
goals, you know, quite frankly made everybody work in the best interest of
the product.
Although the above mentioned goal alignment is often used to explain why teams work,
it is not enough to guarantee close integration within a team. It is possible for different
team members to share a goal while simultaneously maintaining their particular
perspective on what that goal means. The difference in this case is the scope and nature
of the goal that the team set for itself. On several occasions, the informants made it clear
that their goal was not simply to produce a product that would be desirable in the
marketplace. but one that would be a financial success as well. One informant described
this as a critical success factor" that is "really underlooked" and "not well-understood."
This broad definition of the goal--the focus on financial success-must
have played a
key role in opening up the space necessary for the close working relationship to develop
between finance and engineering. By focusing on both market as well as financial
success, the finance people on the team had to develop an appreciation for what the
market was asking for in terms of features and performance characteristics. They could
no longer simply be "the people who said 'No. It costs too much."' Similarly, the
engineers had to evolve the same concern for cost and profit that previously had been the
province of the "bean-counters."
Within this atmosphere, where every team member was concerned with the other
members' problems, compromise was not a zero-sum game. According to one informant:
"Everybody's needs were supported sufficiently so that we had a good balance of
product." The free flow of information between team members and a focus on
communication are prerequisites for this type of open atmosphere. This was described by
the Executive Engineer in the following comment:
Because there was more conversation from more different points of view
to basically create, not a consensus necessarily, but to create a nice
130
picture of'[the situation.]Ifyou were looking at a feature discussionor a
fourth door, so what do you see good about it, what do you totally see bad
about it, notjustfrom a technicalperspective,butfrom a marketing
perspective,froma designperspective,stylingperspective. Youcan see all
qf that conversationcome togetherat one time, the manufacturingside of
it too. And, quitefrankly, itjust sort of helpedlevel the playingfield of
information that everybody that was relevant to making the decision had.
Another indication that the minivan team was a truly integrated team is that team
members were not afraid to admit their failures to each other, and to ask for help, as
explained by the engineering manager in charge of noise and vibration (NVH):
But I think./br me personally, the team also provided the environment
conducive to self catharsis. I was no longer qfraid to say I have failed
someplace, because I knew somebody was going to help me out. As
opposed to bejbre, How am I going to say that? Somebody is going to nail
me.
At this point, it is useful to discuss some of the factors that contributed to the success of
the minivan team. A key "enabler", mentioned by the informants on several occasions,
was the "enormously supportive" upper management team that was in place at Chrysler.
Equally important, this support came with relatively little meddling in the team's decision
making. Again, the NVH manager:
Management this time was a partner. as opposed to previous years [...
when] we were always aftaid This time they were our.friends. They were
there, and they supportedanythingthat really made sense andyou know,
they really turned us loose to do the engineering, to do the development, to
do the selling, to do the advertising, to do everything.
An important factor must have been the deep sense of responsibility that the team
members felt towards the rest of the corpora,;on, due to the importance of the minivan to
Chrysler. Presumably, the feeling that the "survival" of the corporation was at stake must
have played a key role in focusing every member's attention on the financial success of
the product as the ultimate, overarching goal.
The team members were also cognizant that they were "carrying a heavy tradition", in
view of the success of Chrysler's previous minivans. Accordingly, the team members
131
were "extremely motivated [and] passionate," and they "were always charged up,
continuously." Their pride in their accomplishment was clearly apparent throughout the
interview.
Achieving such a tightly integrated team was clearly helped by the fact that many team
members were also minivan users (see previous comment about "having a team that's
used to using these things.") This last point suggests that the distinction made here
between internal integration (within the team) and external integration (with the
customer) is somewhat artificial, that it is more a matter of convenience
than reality. The
engineer's attentiveness to the needs of the customer. which was the subject of the
previous section. may have been, in the end, the result of engineers and marketers
working closely together on the same team.
Conviction,trust, and innovation
The two previous sections described the minivan team's ability to innovate as resulting
from the team members playing various interpretive roles. Yet. the language used most
often by the informants to describe those situations that resulted in innovative ideas was
that of risk, and the team's willingness to take risks. The team members credited the
dedicated team structure with making possible those risky decisions that relied more on
gut feel than hard market data. This section attempts to bridge these two aspects of
innovation, namely. risk-taking and interpretation.
First, let's look at the interface between the design team and the customer. The lack of
customer data relating to the decisions in question made them, by definition, risky
decisions. In the analytical language of decision making in uncertain situations. risk is
directly proportional to the information gap. that is. the difference between the available
information and the information that is needed to eliminate any probability of the wrong
decision being made (Suh. 1990). There are several possible responses to such a situation.
The first is to obtain additional data to reduce the risk of the decision to an acceptable
level. A second answer to a situation marked by incomplete information is to choose a
course of action that would be satisfactory, or simply acceptable, under the widest range
of possible outcomes. A third response is to forego making a decision at that point, and to
diversify the risk by pursuing several parallel courses of action, each relating to a
132
different likely outcome. These responses generally entail added expenditures of time and
resources, whether it is the cost of obtaining additional information, or the cost of
simultaneously pursuing more than one track. These added expenditures are often
prohibitive (hence the need for a decision in the first place.) In the case of the second
response, a very important cost may be compromised performance. By seeking to
adequately serve a wide population of potential customers, the product is less likely to
delight any of them. How does the interpretive notion of external integration (or
customer-designer integration) compare to these analytical responses?
In a sense, external integration is about obtaining more information about the customer in
order to make better product decisions. However, the information in question does not
necessarily fit the definition of hard analytical data, as envisaged by the analytical model.
Rather, the "data" might take the form of tacit elements in a vision of the product that is
yet to be fully understood or articulated, let alone coded in precise features and
specifications. Interpretive external integration is about reaching a stage in which the
designer, by sharing enough of the relevant aspects of the life of the customer, gains a
kind of implicit understanding of the needs, wants, and concerns of that customer, to the
point where such a vision can begin to take shape in his or her mind. This is quite
different from gathering "analytical" data through clinics or test markets.
The analytical thinking about risk management and the interpretive approach differ
drastically when it comes to the second possible approach to managing risk, namely,
choosing a compromise solution. One obvious compromise answer to the fourth door
dilemma would have been for the team to design a normal hinged door. It would have
been less expensive to engineer and build, and since an open hinged door is easily visible
to drivers approaching from behind, it would have allayed the safety concerns brought up
by those who saw the prototypes. The downside of such a design would have been in the
areas of convenience and utility. That compromise solution was considered and rejected
by the team. Instead of compromise, interpretive thinking is more likely to lead to a sense
of conviction emerging within the design team, conviction that stems from their implicit
or visceral understanding of who the customer is and what she wants, and that leads them
133
to see one particular solution is the best one for that customer. That sense of conviction is
captured in the following by one of the informants:
Quite frankly, we trusted our gut because, if you go hack right through the
numberof iterationsofdfferent styles of door-we lookedat hinged and
sliding--and [...] we kept trying to convince ourselves to do a hinged door
[hbecause]it would he ea.sy.But [...] you kept circling hack around it, and
every time you looked at it, you said, f I'm going to do it, I ought to do it
as a slider. It was exactly the right thing to do because it had the most
utility as a slider. And. ultimately.we made that decision,okay,and I think
we, quite frankly, in making that decision we set the standard.for the
industryagain [...]
A compromise solution or half-measure will generally show up in the product and will be
readily apparent to the customers and the critics when they are first exposed to the
product. especially since they are not privy to all the trade-offs inherent in the decision.
The designers, on the other hand. would have been living with the product for some time
before its release to the market, and they would have come to accept the decision. and
come to see it as the best answer to the well-circumscribed cost-benefit problem in
question. From an analytical point of view. given a certain objective function combining
utility, safety and cost. the hinged door solution might have been optimal. How a minivan
with a sliding door on one side and a hinged door on the other would have been received
by the public is less clear. At best, the lack of symmetry would have raised some
questions. The team's conviction in this case not only resulted in a product with a unified
design aesthetic and a sense of integrity, but their interpretive approach to the problem
allowed them to see beyond the parameters that were germane to the trade-off, and to see
possibilities and synergies in areas that are only indirectly connected to the decision
itself. This is clear from the following comment:
And, then an extension of that was the decision to project that we would
put more of the captain'schairs in thesecond [rou, insteadof a bench
seat across the back and what evolved was, now you've got.fbur people,
each with their own door and their own seat[...] So it offers something
that a car doesn'tin thatyou've got all this individuality.You'vegot two
teenage kids that hate each other; they don't want to be close to each
other. They sit next to each other, but they [...] are separate; separate
seats, separate doors, separate everything.
134
Finally, as far as the third approach to risk management is concerned, namely,
diversifying risk by pursuing different parallel tracks, the story of the minivan team and
the fourth door decision does not have much to teach us. Since the fourth door is offered
as an option, not as standard equipment on all minivans produced, one could say that it
conforms to that model of risk diversification. It is therefore difficult to draw distinctions
from that example concerning any differences between the analytical and the interpretive
approaches regarding this approach.
Now, let's turn to issues of internal integration among the different team members. From
this perspective as well, risky decisions abound: at any point in time during the
development process, the team has to rely on information from team members
representing the various technical areas, information that is generally incomplete,
provisional and equivocal. In addition, the actors may have incentives to conceal and
distort information. For example. it may be that preliminary noise and vibration test data
do not meet the expectations or the goals set by the development team, and the NVH
development engineer may decide not to share that information, in the hope that his
group will be able to remedy the situation in the near future. Or, in the case of packaging.
the powertrain designers may intentionally ask for more space in the engine bay than they
really need, just to be on the safe side, in case cooling problems arise for example, and to
make their job easier. Given all this uncertainty, the traditional responses to managing
risk are the same ones described above: expend more time and effort to get better
information; compromise the design to allow for a range of outcomes; diversify the risk
by following several parallel solutions. In all cases, the downside is increased cost.
delays, and a loss of integrity in the result. The Chrysler minivan team appears to have
avoided these pitfalls, thanks to the tight integration within the team and the resulting
high level of trust between the team members.
During the interviews, openness and trust emerged as key aspects of internal integration.
playing a role equivalent to that played by conviction in the case of external integration.
In a dedicated platform team, the team members are more comfortable making risky
decisions, because they perceive the quality of the information available to them as being
greater. At a first level, this is due to the increased trust that team members have in each
135
others' technical competence that results from them working closely together for a
significant period of time on a single project. Here is how one informant put it:
There 's a cornfort in that you know that the other guy... For instance, if
you 're the marketing guy and I'm the engineering, I'm not necessarily sure
you know what the hell you're doing, right? But f you're working with her
every day, you say, the marketing guys don't know a lot about engineering,
but it looks like they know a lot about marketing. So I can trust you a little
more. So the level of trust goes up and therefbre the level of confidence
goes tip.
At a higher level, the trust among team members shifts from a trust in the other members'
technical abilities, to a belief that all the team members are going to place the common
good of the whole team ahead of their personal or parochial interests. The following
quote makes it clear that such a level of trust is present within the Minivan team,
although, in this case. the informant attributes it to the instrumentality of goal alignment
among the team members:
[In the old days, you probably wouldn't have a meeting with all those
people around the same table unlessyou had a crisis of enormous
importance.]Or, f you did you weren'tsure that they were all going to
play their hand out the first time. I think that's one of the key
differences... And it goes hack to beingfinancially succe.ssfulas a
platfbrm. Whenyou're committedto beingfinanciallysuccessfulas a
platform, then nobody has a reason to hide anything or delay saying
something [There are] no ambushes.
This increased level of trust between team members translated into a higher level of
comfort with taking risky decisions. as the following quote indicates:
I think it's comfort with the risk. You can be more comfortable at a higher
level of risk becauseyou understandmoreabout it, instead of the classic
way of doing it, marketingfires a salvo at engineering,and engineering
fires a salvo at manufacturing,and theyfire a salvo over at sales, and
everybodygoes aroundsaying, I don't know if I believeyou [...]
At the next level, trust affects more than the team members' perception of the quality of
the information that they are getting from one another. As the level of openness and trust
within the team increases, team members feel comfortable taking more chances and
136
trying more innovative solution to their problems, because the fear of failure (and
retribution in case of failure) is reduced, as we saw earlier.
A dedicated team with so much conviction in its own beliefs and so much trust between
its members is like a high-gain closed-loop system: it is capable of very high
performance, but it could also easily go unstable. What prevented the minivan team
members from losing touch with reality by creating a distorted version of their own, and
from trusting and supporting each other all the way over the deep end? One form of
control came from the team members themselves. The trust between the team members
was neither blind nor unconditional. The core team members were experienced in
automobile design and carefully scrutinized each other's decisions:
Everybody was always over cost, and you always had people kind of
nippingat you, and challengingwhatyou were doing.All constructively.
And:
Being a team doesn't mean that this is a love-in. It's like a.family and
everybody's got their own little part in it, and so they all challenge each
other.
The other source of control was external to the team, as upper management provided a
form of reality check. Generally, all the informants praised upper management a number
of times for being supportive of the team, and for trusting the team and allowing it to do
the job it was assigned without meddling. Following are two typical comments:
Managementin this time was a partner as opposedto previousyears that I
was there, andfor good reason, that they... we were always qfraid. This
time they were ourfriends. They were there, and they supported anything
that really made sense and, you know, they really turned us loose to do the
engineering, to do the development, to do the selling, to do the advertising,
to do everything. And that is really quite different when management is
reallypart ofyour team and a good partner:you have added confidence
in the importanceof the project [...j
And:
And, in the past we were more driven by [...] what the top guy wanted, on
a given Tuesday.
137
However, in at least one instance, upper management played an important role in
redirecting the efforts of the team. That was in the areas of noise and vibration (NVH),
and it happened late in the development process, when it became clear that :!e prototypes
were falling short of the expected performance levels, as described by one informant:
Let's talk aboutthe total NVHpicture. Wepresented let's say, a status
report to management, through ride-and-drive, and they were not happy
with where the NVH was. That was almost late in theprogram. [...] So,
what happened was, we went to work from the technical side, so we put
togethera propertythat says here'show we ought to be.
It should be pointed out however that the team's esprit de corps and the supervisory role
of upper management did not translate into an "us versus them" attitude. The interviews
indicate that there was mutual respect and trust between the core team and upper
management, who were seen by the team members as equally competent and passionate
about the product. This is captured in the following comment:
It was very comforting when you [...] have your high-level management
people, andproduct planners,marketing,and the only thingyou could
hear, I remember this specific instance, I think, it was. GO FOR IT! To
me, the president and the vice-chairman, [when he gives credence to your
gutfeel, I mean there's got to he somethingright. [...] Now you 're
approaching the area that we 're going to talk about, the passion that we
have.for the automobile.
Without the strong cohesiveness and trust within the team, and without the strong support
of upper management, it is difficult to imagine how the team members would have felt
confident enough in their judgment to pursue directions that clearly ran against what the
analytical market research findings were telling them.
The Design Office
The dedicated platform team is one of the most significant recent developments in the
organization of new product development at Chrysler. Still, the most visible element of
the new minivan was the work of the staff of an "old-style", functionally specialized
corporate office, the Design Office (DO). In view of the importance of design to the
success of several recent Chrysler products, one of two directors in the Design Office was
138
interviewed at length, and the main findings are presented in this section. To summarize,
these are three. The first is a strong connection between innovation and creativity on the
one hand, and the relaxation of practical constraints on the other. The second is the
responsibility of the designer in leading the customer, a different relationship than the
customer-designer relationship that emerged from the NS Minivan team interviews, and
the strong fashion aspect of design that is behind this attitude. The third finding relates to
the sources of new ideas.
The Design Office is a corporate office responsible for exterior design (previously
referred to as "styling"), interior design, fabrics, and colors. It employs about 330 people,
and is organized into two groups under two directors. One is responsible for the exterior
design of passenger cars, including minivans. The other is in charge of designing jeeps
and trucks, and is responsible for interior design, fabrics, and colors on all Chrysler Corp.
vehicles. At the time of the development of the NS minivan, the Design Office
directorates were further divided into an Advanced Packaging Group and a Production
Group. Advanced Packaging works on the styling and general layout (or packaging) of
the vehicle in the early design or concept phase. The Production Group gets involved in
the next phase of the process, in order to refine the design and make it production ready.
The DO also comprises an ancillary group in California called Pacifica Group, whose
task is to provide alternative ideas and concepts. Pacifica's work is limited to advanced
packaging, as they have no production design capability.
The corporate Design Office is not part of any platform, but they work with all the
platform teams. In the new dedicated platform team organization, the team itself--in
particular the DO staff members who are quasi-permanently assigned to the team--would
start thinking about the next generation model and developing ideas and proposals for it
as soon as their work on the current one is done. In the case of the NS minivan product,
the platform organization was still being developed when the project was launched, and
the work on the various design proposals therefore came directly from the DO. The
details follow.
139
Creativityv. practical considerations
The Advanced Packaging Group and Pacifica initially developed a total of four
proposals. Their work was described by our informant as "very right brain", "very
creative , not all that analytical". These proposals were based on general organizing
themes, reflected in the names they were given: "Silver Bullet", "Reach", "Response",
and "Euro". In the case of "silver bullet", the studio was told, "if you had one shot to do
the best next-generation minivan, based on everything you know, all the data that you can
get, all the experience that you've had, and seeing them and driving them and so on, what
would that next generation minivan be like?" "Reach" on the other hand was based on the
idea that, given that the "silver bullet" design was the next minivan, what would the
generation after that look like?
In the case of' "reach" designs. the designers are told to be less constrained by practical
considerations. They are encouraged to make optimistic assumptions about technological
advances that would affect their work. For example. a designer may choose to assume
that there is going to be a breakthrough in the cooling system of the engine, allowing a
significantly smaller front overhang (the distance between the forward edge of the front
bumper and the front axle.) The relaxation of existing practical constraints is seen as an
important ga:eway to creativity, as our informant put it:
When it's a 'reach'property. it's a little bit.further out, thingscan befar
more creative, you can do things that... it might be a crazy idea, but
maybe it will work, maybe it will look good, maybe it won 't look good.
Maybe it will be functional, maybe it won't be.
"Response", the third preliminary proposal, was the result of interactions between the DO
and the Executive Engineer in charge of body engineering, a core member of the NS
team, which was in the process of being assembled at that time. "Response" was basically
"silver bullet" modified to allow for more practical engineering considerations and
constraints, such as structural members of realistic dimensions.
Pacifica:creativitythoughisolation
The fourth proposal was done by Chrysler's Pacifica group, which was tasked with doing
what they thought would be the "ideal European minivan". From the interview material.
140
it is not clear why Pacifica was given that particular task, or what the significance of
Pacifica's contribution to the overall project was. Still, Pacifica illustrates very well the
underlying notion that permeated the whole interview, namely, that creativity can only
blossom when practical considerations and constraints are relaxed and even disregarded.
Pacifica group consists of approximately 20 people, located in California. Their task is to
provide the corporation with an alternative viewpoint and product ideas. To that end,
visits to Pacifica by personnel or management from Detroit are intentionally kept to a
minimum. For instance, major corporate visits to Pacifica take place only about once a
year. The reasoning is that Pacifica's personnel must be kept away from the daily
concerns of their counterparts in Detroit in order for their thinking to remain independent,
and their ideas different. As our informant put it:
Pacificawe generallyuse to give us alternativepoints of viewthat are not
constrained by... I mean, we don't go out there that much. We have
corporate visits out there, major corporate visits, maybe once a year. But
we intentionally keep them away from a lot of the day to day kind of
business that we have here, with the intent that they'll come up with
somethingthat we aren'tthinkingabout becausethey're not constrained
by the consistent thought process that's running through the offices here.
Later in the interview, he sharpened the conflict between creativity and day-to-day
concerns. From the following comment however, it is not clear whether the Pacifica
designers are expected to come up with new and different solutions to given problems
because they are removed from the dominant way of thinking of the rest of the
corporation, or whether they are simply being asked to develop ideas on the assumption
or the wish that these problems do not exist in the first place. or that they are somehow
solvable:
We try to maximize their [Pacifica] creativity, minimize the influences that
would tend to homogenize it and bring it down to the level of, 'But don't
you know that wouldbe hard to manufacture?' 'Don'tyou know it would
be hard to package it?' Or, 'Don'tyou know that the coolingproblems
that we were having with whatever? '[...]
141
This notion of helping creativity through isolation from the daily problems and concerns
finds its ultimate expression in the "sabbatical room" at Pacifica. The purpose of this
room was explained as follows:
At Pacifica,.for example, we have a sabbatical room. You know, the theory
behind the sabbatical room is that a designer can schedule a week off like
he does vacation, except he doesn't [go on vacation]; he's still at work;
he's just in a different room, and he's not involved with any of the projects
over here. And what he's expected to do is to do something we haven't
seen before. [...] And, the intent, again: Here is the studio that is most
removedfrom Chrysler to begin with, [and we 're] saying that even within
that environment, designers need to have some time when they're not just
working on something that their management says that they should work
on. So fine, go out into this room; you've got a week show me something
that I haven't seen before. And.from that sort of thing, we would expect to
get thingsthat are, whatever'sgoing to come aqfter'Heritage'[Chrysler's
current design trend], and whatever's going to come after Ford's 'Edge'
[Ford's latest designtheme], or whatever.
It is noteworthy that Pacifica's immersion in the Southern California car culture was
never mentioned in the interview. When US companies set up satellite design offices on
the West Coast in the 1980s, one often-mentioned rationale was the need to tap into the
California car culture, which is considered a harbinger of the trends that will come later
to the rest of the country. It may be that this particular rationale for Pacifica's existence is
so entrenched that it was taken for granted and never mentioned. More likely, it seems
that Pacifica's significance and importance lays more in its distance away from Detroit
than in its location in California.
Design decisions:conviction,responsibility,and what the boss
wants
We now turn to the process of selecting from among the different proposals described
above. The focus in this section is on the role of the customer, the role of the designers,
and the role of upper management in this selection process.
The four proposals, "Silver bullet". "Reach"', "Response", and "Euro", were developed
into full-size fiberglass mockups, complete with representative exteriors, interiors, and
seating areas. These mockups were used in customer research clinics, in which
142
participants were asked about the desirability of certain features, such as the fourth door,
and to rate the different proposals using a list of 25 descriptive terms such as "sporty",
"luxurious", and "efficient". The data were then sorted based on the informants' age, sex,
income, and so on. From experience, Chrysler knows that certain cohorts are better
predictors of future trends in certain product areas than others, and the information was
used to "give us [the designers] the best example of what direction we should take the
vehicle in."
It is not clear from the interview how closely these customer data were really used. The
informant went on to say that sometimes, such data is ignored, as the designers relied on
their own conviction of where the product design should go in the future. The reason
given for ignoring customer input is the time gap between the future life of the customer
into which the new product is intended to fit, and the current world in which the customer
lives.
... sometimesyou have to ignoreit all, becauseyou recognizethat all of
thesepeople are living in that currentday;you're planning on something
three years awayfrom that day, and sometimes you have to have the
courage ofyour own convictionsto say that, I know that threeyearsfrom
now...
The designers on the other hand, by the nature of the work they do, are analyzing data
about the future and studying trends, they are thinking about the future and imagining
what it could be like. They are effectively living in the future, and are therefore in a better
position to "know" what people will want and need a few years down the line.
The above explanation about the time lag inherent in working on future projects could be
taken at face value, and it would make sense from an analytical perspective. If there is a
preexisting future reality out there that is coming at some point in time, then it would
only make sense that people who are actively involved in tracking trends and analyzing
data on demographics would be in a better position to see that future, compared to the
average customer who is concentrating on the daily concerns of today. However, there is
another dimension to the informant's explanation that is not captured by analytical
thinking. Although it is not articulated during the interview, it is hinted at in a number of
the responses and comments, including the one above, in the form of the "courage of [the
143
designer's] convictions" is mentioned above. This additional dimension also shows up as
a responsibility on the part of the designer, which is not merely about getting a preexisting truth from the customer about the future, but goes beyond to contribute in
shaping that future. The example given to illustrate the point was the introduction of the
original minivan in the 1980s. The research data on that design was negative, as many
informants "thought it was homely, and could not see themselves buying one."
The special responsibility of the designer became clear as the informant went on to
describe the next phase in the design process. The four concept proposals are used to get
"concept approval", a point at which all the parameters of the package are set, including
cost, projected volume, engine choice, suspension design, and so on. At that point, the
designers go into "execution mode", and develop a number of "themes" based on that
package definition and the now fixed dimensions with which they have to work.
Generally, between six and ten themes will be developed in clay. and about four in
fiberglass. These are used in the "theme selection" process. Here again, the role of
customer input into the decision making process is even less clear. if there is any. Asked
if any consumer clinics are used for theme selection, our informant answered "We clinic
them after we've chosen them" and laughed. He went on to explain
It's notfair... It s not fair to ask the customer... To me, it's a real executive
cop-out to say, well, the research customer told me this was the best
theme, so I guess that must be it. [...] It's not fair to ask them to make the
decision. It's not fair to your own corporation to say, you know, I'm only
here to run the research clinics. My gosh. You 've been hired to be the
expert in design; why not be the expert in design?
Although the informant did not further articulate what it means to be "the expert in
design", it is clear that his understanding of that responsibility went beyond mechanically
translating the customer's pre-existing preferences. In fact, one can sense a hint of a
patrician view of the designer's role, one that puts the designer in a privileged if not allknowing position. This special role of the designer showed up in a later comment, as the
informant was further describing the theme selection process. This is a multi-step
process. In the first step, the designers vote on their favorite theme; but it is only the
design professionals, people who understand design, who get to vote:
144
The process we go through right now is, we'll put the models in our
showroom; we'll have everybody from the design office come in and vote
on it. From the design office! These are all the designers and the modelers
and, you know, people who are sensitive to design, trained in design. We'll
have them come in and vote. We've ignored that vote in the past.
Still, the professional designers are not the ultimate decision makers, and their vote is
sometimes ignored by the design executives. Admittedly, this has only happened in rare
cases, where the voting was very close. Still, one is left to wonder about who the ultimate
decision makers are. Our informant explained that the styling decisions are really made
by the top executives of the corporation, with recommendations from the different levels
of executives:
Realistically, the decision is confirmed by the broader group; the decision
is made by the design office with our vice president, John Hurlitz, with our
executive vice president who is also in charge oJfproduct development,
Tom Gale; by the president of the company, who is now Bob Eaton, and by
our head coach, Bob Lutz. I mean, that's reallythe group that makes that
decision.
This was later confirmed by another informant, a director in Advanced Manufacturing
Engineering, as she commented on how her group might support design and styling
decisions:
Well, that's not how decisions are necessarily made on aesthetics. In our
company it's more of a collaboration of the top officers in the company
who have a tremendousbase of experiencein that area, and of the design
office bringing forward choices. And what we'll have to work with is to
make sure that when a team ofpeople are making decisionsabout whether
a looks better than b -- and much of that is purely subjectiverather than
objective-- we can help them to understand,you know,that here is a and
it costsx and here is b and it costsy. And they can then makea more
informedbusinessdecision, so it's notpurely an aestheticdecision.
The fact that the top executives at Chrysler appear to reserve to themselves such a
decision making role is at odds with current management thinking, where the push is to
move decision making as low as possible within the corporate hierarchy, and to rely as
much as possible on customer input in designing new products. It does however explain
145
an expression that was mentioned several times during the interview, about the "courage
of the executive conviction."
The notion that conviction on the part of upper management is an important element in
the way styling decisions are arrived at is closely related to how they see the role of the
corporation in the marketplace. Chrysler management often talk about wanting to design
cars that people can be passionate about, and they feel that passion in the customer cannot
be aroused, and the consumer drawn to the product, by producing designs that appeal to
the greatest common denominator of consumer preferences. There are many examples of
such designs in Chrysler's portfolio, such as the Dodge Viper and the Plymouth Prowler,
which are low-volume specialized vehicles. But the same thinking is present in highvolume high-margins products, such as the very popular Ram trucks. The story of that
design follows:
[...] when I say we want to build cars and trucksthatpeople are
passionate about, it says that we are willing to do things that are more
polarizing, we're willing to do things that don't.fit the average description.
I'll use the Ram truck as an example. When we were doing [...] the
packaging on the Ram truck [...] we were trying a lot of different things,
and we were researchingthem, and if it didn't look like a Ford truck or a
Chevy truck it was, ff..ft, forget it. It had to look like one of those, or
forget it, nobody wanted it. So, ifyou were to listen to research, you would
do a new truck-because this was going to be ourfirst new truck in 20
years or whatever-you woulddo one that lookedjust like a Ford or a
Chevy, and that would have failed, or we probably would have kept our 7
or 8% of the market.But that would havefailed, becauseit wasn't
anythinganybodycould be passionateabout. Theyalreadyowned Fords
and Chevys and why would they buy a Dodge when they could have the
real thing? [...] So insteadof doing that, we did one that was polarizing,
that when we took it out and researched it, there were some people that
loved it [his emphasis],somepeople that hated it [ditto].,and but it wasn't
in the middle.And that's the kind of reactionthat we want to have to our
products. We want to have some people love it, be passionate about it:
and it doesn'tmatter f therest of the people aren't. It comes back to [the
.factthat]people never buytheir second choicevehicle. "It s really nice"
but they don't buy their secondchoice, they buy theirfirst choice:
assuming they can afford it, I mean.
146
Why have a separate DO?
The answer to this question is twofold. First, the DO is responsible for developing new
ideas that would not fit, at least not a priori, within the mission of any particular team.
Second, the DO performs an integrative function across teams as it disseminates some of
these ideas. The following sections discuss these two roles in detail.
1. Developing and exploring new ideas
Within this role, one can distinguish three classes of activities. In the first category, the
activities are product or platform specific. The best example of such activity is the work
done by the Design Office for the minivan platform early on, when they developed a
range of studies as described earlier. Although platform specific, this work went beyond
what would have been done within the platform team itself. The early minivan studies
were described by our informant as having "corporate visibility" or "corporate focus",
meaning that the top management at Chrysler was aware of and specifically approved the
range of ideas covered by these studies. They felt that the range of ideas was broad
enough that these studies would contribute to the future design direction of the minivan
and of the corporation as a whole. Our informant explained that, had the Design Office
been purely responsible to the minivan team, they would have only developed one of
these studies, specifically, the "Response" study. Given the time and profitability
constraints under which the team must operate, it would have focused on the readily
feasible and immediately relevant, and it would have found it difficult to fund such a
wide range of studies. Our informant tied the independence of the DO to its ability to be
more creative and less constrained by accepted thinking and market data. This is all
captured in the following quote:
These [the different minivan studies-'Silver bullet', 'Reach '. 'Response',
and 'Euro 7 were all done about the same time, and these were done about
three and a half tofour years ahead of launch.So, then we went out and
researchedthose,and we were talking,as far as visibilityis concerned,
corporatevisibility;and the presidentand so on are all aware of what
we're doing: 'Yeah,that's right, that's a good spreadof ways of looking at
the vehicleandfrom this we think we canget a good answer.' If we had
had strictly teamresponsibility,in other words, if theplatformteam
dictated what we would do, the 'Response 'may have been the only vehicle
we would do, because that was in response to the team. So, by being
147
independent, we have some greater ability to go out of the box in our
thinkingand try to reachfor decisionsthat are morecreative,that don't fit
the 'analysis'.
The cost to the corporation of not having this range of design exploration work performed
goes beyond the danger of ending up with a bland, "me-too" minivan in this product
cycle. The greater danger lies in losing important elements of conversation within the
organization, between engineering and styling for example, relating to future product
cycles. By proposing a futuristic "Reach" design that requires some super-efficient not
yet developed cooling technology, Pacifica might push the cooling system people within
Chrysler or within its family of suppliers to pursue a product development direction that
they would not normally have followed, and thus to make such a design feasible at a
future point in time.
The second class of activity in which the DO engages, and which is unlikely to be
undertaken by the platform teams. is the work of developing corporate or brand design
themes. This type of work was touched upon earlier, in the context of describing the
sabbatical room at Pacifica, and it was described by our informant as being "a matter of
continual experimentation and investigation and, you know, [asking ourselves] What else
can we do..." Developing new corporate design trends involves a long lead-time, and a
significant amount of trial and error. In addition. these trends change continually:
[...] heritagedesign or cabforwardare specificsthat we'reworkingon
now, thatfifteen yearsfiom now, or tenyears.from now, we'll probably be
working on something else. So they're not the sort of thing that you can
say, well here [it] is...
For example, in the early 1990s Chrysler's main design trend was the so-called '"cabforward" design, where the passenger compartment was stretched forward as far as
possible, resulting in the distinctive look of the full-size LH sedans. Cab-forward, as a
design trend. initially began with a design done at Pacifica called the "Navajo". which
was a 'reach" study for the LH sedans of the early
990s. something the LH platform
team would probably not have requested. The objective of the study was to reach beyond
what was conventional in order to give the passengers of the sedan more interior space.
The "Navajo" study was later transformed into a mid-engine sedan called the
148
"Portofino", after Chrysler purchased Lamborghini, a small Italian manufacturer of highperformance mid-engine sports cars. The mid-engine configuration of"Portofino"
accentuated its cab-forward configuration proportions. As a show car, "Portofino" met
with such success that the CEO of Chrysler at the time, Lee lacocca, became a convert to
cab-forward, after initially being against it. The time and resources required to refine a
new design theme over several generations of concept cars, and to prepare the different
stakeholders within the organization to accept the new theme, are simply not available to
a platform team. 4 5
More recently. Chrysler has embraced the "heritage" design trend that has been sweeping
the industry and that has resulted in such models as the Porsche Boxster and the BMW
Z3. Tapping into these industry trends and developing the next ones requires people who
are in tune with what is going on in their community of practice outside of the
organization for which they work. At the same time, it requires a community of designers
of a certain critical mass within the organization working together, exchanging,
discussing and interpreting ideas and trends (Seely Brown & Duguid, 1991). That is what
the DO provides.
The third category of creative activity that is the exclusive province of the Design Office
is the creation of totally new products that do not fit, a priori, within the mission of the
existing platform teams. Such products would not simply be extensions of existing
models, such as a longer wheelbase or a more luxurious version being added to the
minivan lineup, or a stretched cab pickup truck being added to the truck lineup. The best
example of such products are the Dodge Viper and the Plymouth Prowler. The Prowler.
which was in the process of being launched at the time of this interview. is a modem
4'
Another good exampleof the lengthy and complicated gestation of design trends is the current trend at
Ford, referred to as "edge design". Initially, Ford designers started experimenting with basic geometric
shapes as the governing theme. This resulted in the current Taurus, with its multiplicity of oval shaped
features, from 'he central console to the rear glass. The oval was a natural shape for Ford whose logo is
the blue oval. Following the same basic idea of using one geometric form for as many elements of the car
design as possible, Ford developed the GT90 concept car. a mid-engine sports-racing car intended to be
the spiritual successor of Ford's GT-40 Le Mans winner of the 1960s. The GT90 used triangular shapes
as the basic design element, resulting in triangular exterior panels, air intakes, and interior design
elements. Elements of the two themes found their way to various production cars; the Ford Ka, a minisize model available only in Europe being an example (ovoid silhouette, headlights with sharp edges.)
The combination of the two design trends, ovals and triangles, became Ford's design theme, and was
149
interpretation of the hot-rod. It is a two-seat open car, with motorcycle-style
front fenders
(the front fenders are not part of the bodywork; they are attached to the front suspension
and turn with the front wheels.) The story of the gestation of the Prowler, as recounted by
our informant, is instructive. The Prowler started with a brainstorming exercise at
Pacifica, an idea fair, where everybody, designers, modelers, even the janitor, was invited
to come up with an idea for a transportation vehicle, no matter how unusual or off-thewall, and to present it. One of the designers came up with the idea of doing a hot rod. He
was encouraged to do a few sketches of his idea. The design managers liked the sketches
and decided that building a few scale models would be worthwhile. This resulted in a
total of five dioramas being built. (A diorama is a scale model sitting in front of a blowup picture of some scene as background, with a few elements in the foreground, intended
to show the idea in a somewhat realistic setting.) These scale models were seen by Bob
Lutz, Chrysler's president. and Tom Gale, then VP of design. As it happens, Tom Gale is
himself very interested in hot-rodding. They found the idea interesting, and gave their
approval for a full-scale clay model to be built, which v, as followed by a show car being
approved. The response by the public and the press to the show car was so overwhelming
that Chrysler management decided to produce it as a relatively high-priced specialty
model, in limited numbers like the Dodge Viper. The decision was buttressed by
marketing and engineering considerations. From the marketing viewpoint, the Prowler
would give the lackluster Plymouth brand a "shot in the arm". From the technical point of
view, the Prowler would be an opportunity to gain valuable design and manufacturing
know-how in aluminum-bodied vehicles, with relatively little risk. The important aspect
of the story, however, is that Prowler came about by the Design Office providing "an
opportunity for a designer just to do something that he felt passionate about, something
that [...] turns [him] on, and that's a hot button... [and to] get other people passionate
about it..."
2. The DO's integrative role
The corporate Design Office plays a crucial role when it comes to expressing the
distinctive identities or character of the various Chrysler brands. In the previous section,
further refined in a number of studies for the upcoming Ford Escort. The recently introduced Mercury
Cougar is the first production model officially referred to as an "edge" design.
150
the development of corporate design themes by the DO was discussed in the context of
the long time horizons involved, and the disadvantageous cost-benefit calculus that would
prevent any particular platform team from undertaking that type of design activity. In this
section, we delve deeper into the interpretive nature of such work, in order to understand
the communication and integration role performed by the DO.
A company needs to have a corporate identity in order to differentiate itself from its
competition. In the automobile industry, brand identity and product design or styling have
been very closely linked for several decades. At Chrysler, the challenge is to give each of
the four brands, Chrysler, Dodge, Plymouth, and Jeep, its own design identity. The
executives responsible for the different brands want their product offerings to have a
recognizable look that is, at the same time, distinct from that of the other brands. Our
informant put it as follows:
The brandmanagers,the marketingpeople, want to have a consistent
look They're concerned with their showroom; when they look at their
showroom and see all the Chryslers, that they want them all to look like
Chryslers.
These executives want their products to convey a consistent message. Generally,
marketing will develop a descriptive message for the particular brand, based on where
they wish to position the brand relative to its competition. That message however will
need to be interpreted into a design. Referring to the Dodge brand image, our informant
explained:
Ifyou were to talk to marketingthey'dsay it's 'affordableperformance';
that's the two-word answer. In design we get tofigure out what that looks
like. And all I can say is, a Dodge wants to look muscular, it wants to look
capable, it wants to look strong. [...] And we'll go th. ough these kinds of
things with the designersand, in pretty openconversation,back andforth,
as we're trying to figure out what a dodge looks like and how do you make
a Neon look like a Dodge for example].
Developing a design identity and. more specifically, translating the product positioning
message into a design is something that has to be developed "in the metal", by doing, in
practice; and an important aspect of practice is that it is interpretive: the designer reacts to
what he is doing and to other people's reaction to what he is doing. In this case, the other
151
people would be the other designers in the design office first, the marketing and brand
managers second, and the company's upper management third. The process described is
not unlike that which takes place between an architect and a client: the architect develops
preliminary designs based on his initial understanding of the client's desires; the client
reacts to these designs; the architect in turn reacts to the client's reaction and modifies the
designs, and so on. That is what "open conversation" and "back and forth" in the above
quotation refer to. Even though the designers may be explicitly seeking a design
"formula" for a particular brand's identity. the process remains more evolutionary and the
designs more emergent than formulaic. In response to a question about how the new
front-end look for the Chrysler brand came about, our informant said:
We evolve our way into it, that is the best way I can put it. As we're
designing, we start realizing that we want something different so designers
are suggesting other looks and we finally, maybe don't get the look...
Becauseifyou look at the Sebringgrille, it's a brightcolor,and it's got a
bright bar across the middle. It's not quite the formula. The formula
doesn't have that bright bar across the middle. And, ifyou look at the
Sebring coupe it's got the egg crateand it's got the badge,but it doesn't
have the bright collar, so it's not quite the formula, but we're getting there.
The above description implies that it would be impossible for marketing to develop a
formula for a brand's design identity on paper and then send it out to the different
platform teams for each to interpret and develop to suit their product. Each team's
interpretation would be different, and there would be no guarantee that the results from
the different teams could be fine-tuned or somehow integrated into one family look to
reflect one brand identity. At a minimum. the process would be costly and highly
inefficient, as the interaction described in the previous paragraph would have to be
repeated several times. Furthermore, other practical considerations combine to make the
process even more complex, which casts a greater importance on the central role played
by the DO.
One complicating factor is the asymmetry in the contribution of different platforms to the
design identity of a given brand. Not every brand offers a full range of models based on
all platforms; for example, Chrysler does not offer pickup trucks, nor does it offer a
version of the Neon; and specialty models such as Viper and Prowler are specific to the
152
Dodge brand and the Plymouth brand respectively. Due to this asymmetry, those products
that are specific to the brand, the "stand-alone products", play a disproportionately
greater role in establishing the identity of their brand. In the case of the Dodge brand,
these are the Ram pickup, the Dakota sport-utility vehicle, and Viper. It is through these
stand-alone products that the designers establish the design cues for the Dodge brand. In
designing the products that Dodge shares with other brands, the designers will attempt to
use as many of these design cues as they can, after modifying them to suit the type of
product in question. In the case of the LH sedan, if the process were left to the platform
team, the team would have faced the complicated task of interpreting the Dodge brand
identity, developed elsewhere, for the Dodge version of the LH sedan while at the same
time developing the brand identity for the whole Chrysler range, since the LHS is the
closest thing Chrysler has to a stand-alone product.
The need to re-interpret the design cues of a brand across a varied model lineup makes
the process even less mechanistic. The brand's design identity was described as more of a
"general framework" than a formula, in order to allow design elements developed for one
type of vehicle, say a truck or a sports car, to be re-interpreted for a different type of
vehicle such as a large family sedan. Our informant put it as follows:
But, within that, that look needs to have some flexibility. In the same way
that it does with Dodge.I mean,you can't have thefront ends of all
Dodge'sall lookingidentical.Becausewhatfits on a Ram truck doesn'tfit
on Avenger, so maybe you say the crosshair is the look but it could be a
big brightheader with a brightcrosshairand honeycombgrille textureup
here, or it could be an all body colorfascia with a low mouthand a
crosshair in the middle of that low mouth on a Viper or Avenger. So, it
says that within that generalframework there is going to be variety. And,
Chrysler'sgoing to be the same way. It's in the generalframeworkof the
design that we have. There will be a variety in any given year, and there
will be a variety through time.
Aside from the complication of integrating across non-uniform product lines, there is also
the complication of integrating across time. As discussed previously, design themes
evolve over time in response to a number of factors including the designers' tastes and
experiences, trends in the industry itself as well as in other industries and, generally, in
response to the cultural environment as a whole. Given that fact, and given that different
153
vehicle models are developed at different points in time, the problem of maintaining the
integrity of a brand's design identity in the absence of a central Design Office would be a
difficult one. Our informant, again, describing the difficulty of changing the front-end
look of Chrysler, from the "waterfall" design of the mid- 990s to the current design:
The problem we have, over time, is that images change, hut you don't
changeall Chryslerson the same day, and becauseyou don't do thatyou
end up with... Well, I want to move from this -- because we had the
waterfall,the verticalbargrille, that was the Chryslerlookfir quite a
while -- and I said, well, I really don't... I want to move away from that:
that's not saying what we want to say about Chrysler. Unfbrtunately,you
can't redo all Chryslergrilles on the same day. So, you start doingthem a
piece at a time,with the intentthatfinally you'll walk away out ofyour last
waterfall grille into the thing that you want to have for ('hrysler.
To sum up, the Design Office plays several communicative and integrative roles, or
rather has an integrative role with several facets: A) it is a bridge, in certain areas of
product design, between marketing and individual platform teams; B) it differentiates
between the different brands to give them distinct identities (this can he thought of as the
opposite of integration, but still a function involving similar skills); C) it integrates across
the various models of a given brand and therefore between platform teams; D) it
integrates (or differentiates) across time, as it maintains (or changes) a brand's design
identity across product generations; E) it integrates across the companies boundaries, by
tapping into the overall culture, but more specifically the community of practice of
automotive design as a whole. As discussed above, the type of integration that these roles
call for is not mechanistic but interpretive in nature.
Supplier Relations at Chrysler: Analytical v. Interpretive
Integration
Aside from its daring and innovative styling, Chrysler has also been recognized for its
strong relations with its suppliers. Researchers in management and product development
have been interested in the company because it started relying on outside suppliers in
developing new products earlier than some of its competitors in Detroit. It should be
pointed out, however, that that was not entirely a matter of choice: as the company was
154
rescued from near-bankruptcy in the early 1980s, it was left with reduced in-house
resources, and it emerged as a much less vertically-integrated concern than its
competitors. As a result, Chrysler has had to give greater engineering and development
responsibility to its suppliers, not only for parts but also for increasingly larger
assemblies and subsystems.
One problem with relying on suppliers is that significant portions of the knowledge
needed to design and build the final product no longer reside within the firm but with its
suppliers. This leaves the company potentially vulnerable to a number of dangers: a) the
loss of competitive advantage through the loss of proprietary product and process knowhow; b) the loss of product distinctiveness as the supplier is able to provide its
capabilities to a number of competing firms; c) the transfer of profits from the firm to its
suppliers, as the latter come to own a greater portion of the value-added to the product
(the extreme case is the PC industry, where the profit margins of system builders have
paled compared to those of the firms supplying key components such as Intel and
Microsoft.) Some researchers have argued that Chrysler's role is that of systems
integrator, and it is a role that suppliers would not be able to perform. A car is a very
complex system, and specialized systems knowledge is needed to bring all the
subassemblies together and to manage the interfaces between them to make sure they all
fit and function together as intended. The ability of companies such as Chrysler to
provide that special knowledge and capability, and to maintain it in-house, would
guarantee their continued profitability.4 6 It was therefore surprising to find that Chrysler
had subcontracted the systems integration work for the interior of its future minivan
project to one of its suppliers, Lear Seating. This arrangement, which is described in
detail below, raises important questions about the division of labor between a firm and its
suppliers, and sheds some light on different types of integration tasks.
The supplier in question, Lear, is a $7-billion company. 4 7 Lear has a number of divisions
dedicated to specific customers; for example, the site of this visit is part of the "Chrysler
Division" of the Lear Corporation. This division is a fully self-sustaining business unit,
46For a discussion of the strategic considerations
Whitney, 1996).
involved in such outsourcing decisions, see (Fine &
155
with its own management; its yearly business volume with Chrysler exceeds $500
million. Also present during the visit were staff members from the corporate office of
Lear, who work in support of the different divisions of the corporation. The visit to Lear
dealt mainly with the RS minivan. the next generation model that is still under
development, whereas the focus of the rest of the case study was the current production
model or NS. For the NS, Chrysler performed the interior integration work itself and Lear
simply supplied seats for the European version. In the case of the RS model, Lear is not
doing any work on specific components; it is responsible for the entire task of interior
systems integration for the program. For a minivan. interior and seating is a particularly
important subsystem. Because the new arrangement between Chrysler and Lear involved
a yet-to be-released model, the interviews were rather limited in scope. Nevertheless, it is
possible to briefly but usefully describe the integration and coordination work performed
by Lear's Chrysler Division for the future minivan project.
Lear is responsible for managing all aspects of the RS interior program. This is carried
out from a 20,000 sq. ft. building. mentioned above, dedicated to the entire interior
supply base. At the heart of this operation are two large rooms. the "wall room" and the
"integration room". The "wall room" gets its name from one of its walls which is used to
display the different iterations and changes over time of the design of every interior
component, from its original conceptual sketches to its current status (detail drawings and
samples parts.) One informant estimated that the wall had 17,000 items on it. Lear was
involved from the beginning of the project: two Lear employees were stationed at the
Chrysler Design Office in order to influence the design with their functional and
manufacturing concerns. They served as the liaison between the DO and a core
"creativity team" at Lear, consisting of 13-15 people from engineering, design,
purchasing, manufacturing and service.
The wall is constantly updated by representatives from the various suppliers. Open issues
concerning components or systems are posted on the wall, including build issues relating
to how different pieces interface with one another and how they are fastened together.
The grain, gloss, and color status of each part is tracked and compared to the design
47
Since our visit to Lear, the company has grown closer to a $12b/yr. in sales volume.
156
objective. Similarly for dimensional tolerances, gaps, and fit conditions. Designer and
engineers can place sticky notes on the parts to raise questions or concerns, and those
would be answered by the person responsible for the part in question. Only the person
who put up the note can take it down when he or she is satisfied with the answer or the
resolution of the issue.
As impressive as the "wall room" is, it is but one representation of information stored
through a proprietary software program developed by a Lear's director for the RS
program. Woven into that software program are all the Chrysler official milestones and
PAPs (Part Approval Points). All the milestones to be met, all the different tooling steps,
all tasks that are less than 200 hours are laid out in the program. All tasks are assigned to
people by name. These tasks drive the timing and are used in design reviews and in
management reviews. The program is also used to track the status of fit and finish, color,
grain of the different parts, and all areas of the RS interior program, with the exception of
the financial aspects of suppliers' parts. All this information is available on the network
linking Chrysler, Lear, and all the suppliers. The advantage in keeping the information in
the wall-room however comes through during the meetings when all suppliers come
together to discuss any open issues and any concerns they might have.
Whereas the "wall room" focuses on the product and its various component parts, the
"integration room" deals with the manufacturing process and how these parts will come
together to produce a minivan interior. Laid along the walls of this room are graphical
representations of all the different stations in the assembly plant. Each station in the plant
has one corresponding column on the wall. In that column are listed the parts that go in at
that station, along with details and sketches. This makes it easy to look at issues of DFA
(design for assembly) and to optimize the process as a whole. In general, the push is to
increase the density of the modules to get away from having to assemble a large number
of individual components at the plant. Here again, Lear is on-line with Chrysler: design
reviews can be accessed on-line, and any changes in the manufacturing process or in
other assembly processes can be taken into account and reflected in the "integration
room" displays.
157
Remarkably absent from the interview at Lear was any discussion of gut-feel and risk-
taking, of developing a shared vision for a yet non-existent product, and other such
concepts that were the highlight of the interview with the NS development team. Lear's
presentation was much more of a nuts' and bolts' description of project management
issues-tracking and communicating design changes and parts status in a timely manner,
distributing information to the different project participants, making sure everybody is
aware of what the others are doing and how that might affect their own work, and so on.
The integration tasks that Chrysler has asked to Lear to undertake can be accurately
described as analytical.
The same kind of division of labor, where Chrysler keeps to itself the integration tasks
that we have been describing as interpretive, and where it subcontracts out the integration
work that is essentially mechanistic in nature, can also be seen at the interface with the
customer. For example, both Lear and Chrysler conduct consumer clinics and surveys on
minivan interiors. But the questions asked and the information sought by each are
different. The research conducted by Lear. mostly by the corporate Customer Satisfaction
Group, is intended to understand the ratings achieved by the different vehicles on the
market on J.D. Power's customer surveys. The researchers "dig down" and try to find out
the reasons for what people are saying. A Lear employee from the corporate office gave
the following example:
For instance, if an armrest is rated 'not comfortable'. we try to see why.Is
it too short, too long, too hard? Is it the angle? We look at the
construction, thefoam used, etc. We are interested in continuous
improvement; we talk to OEMs so we won't do redundant clinics and to
share information.
Lear's research can best be described as benchmarking. By contrast, Chrysler's clinics
are intended to measure the reaction of potential customers to different future exterior
and interior designs, different seat packages and trim. A Lear informant described the
difference as follows:
At Lear, our research is much more specific. We look at the whole
segment, so we get a snapshot of the whole minivan market. in other
words, we look at current products. Chrysler is more concerned about
158
how to delight the customer withfuture models: so they usefiberglass
bucks to test new features on future models.
The terms "delight" and "future" capture the essence of the difference between Chrysler's
interpretive work and Lear's more analytical, mechanistic activities.
The findings at Lear Corporation point to an important distinction between different
kinds of integrative activities, the analytical ones on the one hand, and the interpretive
ones on the other. The findings indicate that Chrysler believes that its ability to
successfully perform the interpretive functions, both internally (within the organization)
and externally (at the interface between the organization and its customers), is the key to
its continued competitiveness as an automobile manufacturer. It appears to have no
qualms about handing over the analytical systems integration work to outside suppliers.
159
160
Chapter V: Case Study II
Nissan Design Intemrnational
Nissan Design International (NDI), located in La Jolla in southern California, was
founded in 1979 as an American satellite studio for the Japanese manufacturer. NDI
employs approximately 50 designers, modelers, and engineers. The focus of its work has
been the styling of Nissan vehicles sold primarily in the United States. The company has
also undertaken a number of industrial design projects for outside, non-automotive client
companies. These projects have included designing furniture for children, golf clubs, and
a yacht, among others. NDI is best known however for some very successful Nissan
models, such as the original Pathfinder sport-utility vehicle, the Nissan Quest minivan,
the Infiniti J30, and the Nissan Altima. At the time of the visit to NDI (June 1997), the
second generation Altima was about to be introduced to the US market. The Altima
project became the focus of some of our interviews at NDI, and at Nissan in Japan as well
(see the Nissan Technical Center case study in the next chapter.) A significant part of the
visit was spent interviewing the director of NDI. Jerry Hirshberg, who was initially hired
by Nissan to set up and run the studio (MacKenzie, 1993). Until he was recently
promoted to president, he had been vice-president of NDI, with the president's post being
filled by a Nissan executive sent from Japan to La Jolla for that purpose. Over the years,
NDI and Jerry Hirshberg have become practically synonymous, as he is often interviewed
and quoted directly in industry articles relating to NDI or to new Nissan models; he even
starred in some recent TV advertisements for the company (Rechtin, 1999). At the time
of the interviews, Jerry Hirshberg was in the process of writing a book about his
experiences and his management approach at NDI. That book, entitled "The Creative
Priority", has since been published (Hirshberg, 1998). The other interviewees at NDI
161
were: the studio head and the chief designer who were responsible for the new Altima,
one of the key persons in the Design Context Laboratory, and the person in charge of the
Color Studio.
History of NDI
NDI came about as the result of the vision and actions of Nissan's president in the late
1970s, Takashi Ishihara. Mr. Ishihara, who was described by our informant as a "pretty
visionary guy", and who went on to become Chairman of the Board of Nissan, had by
then come to realize that although the Japanese manufacturer had achieved a very highlevel of technological capability, it was still lagging in styling and design compared to the
top automakers in the world. Unlike many of his peers at the time, Mr. Ishihara was also
cognizant that product design, and in particular the totality of a coi,;pany's designs over a
number of product generations, is a powerful form of communication, a language through
which the identity of a company is shaped in the marketplace. With half of Nissan's
target audience being non-Japanese and non-oriental, he realized that his company was
trying to communicate in a foreign tongue with more than half of its market. He also
realized that, unlike technology and science, something as ingrained in the history and
culture of a society as design would be very difficult to learn and transfer, and it would
have to be dealt with differently.4 8 As our informant put it, "You can't get an injection of
design." Yet, Nissan's president wanted to go "beyond intriguingly well-put-together
products of high quality that were nonetheless quirky and didn't fit the intended market
physically, let alone esthetically."
Nissan could have gone to one of the well-known Italian design houses for some of their
models, as Isuzu did with the Impulse, a sports coupe designed by Giugiaro, or as Toyota
did many years later when it hired Pininfarina to design the original Lexus GS300 sedan.
But it would be impractical and too costly to hire these star design firms to do a whole
product line, over several generations of these products, not to mention the regular
48 Our
informant, Jerry Hirshberg, pointed out that while some third world countries have absorbed
technological innovations from the leading industrial economies at an amazing pace. they have not been
able to come anywhere near the leaders when it comes to design. And those leaders, such as Italy and
now increasingly Spain, countries that continue to struggle in other ways (i.e., economically), have such
a rich cultural heritage in design that design "is in their bones, and they do it beautifully."
162
updates and facelifts that would normally be required over the lifecycle of a model. (It
should be pointed out that certain low-volume, high-end automakers follow just such a
course of action. For example, practically all Ferrari models introduced in the last several
decades have been designed by Pininfarina.) In the absence of such a long-term
committed relationship with one design house, the company would be left with a
disjointed product line-up and with an incoherent message in the marketplace as a result
of the dissonant voices represented by the different products. Mr. Ishihara's answer to the
dilemma Nissan faced was to propose that the company set up its own design studio in
the West.
Five different countries were considered as potential locations for Nissan's new design
studio. These were England, Germany, France, Italy, and the United States, all countries
with "a rich and successful heritage of automobile design". Nissan's president felt that
they could learn about design from any of these countries, and he sent out a team to study
which of these countries should be chosen. Although the Japanese had "great respect for
German and Italian design," and a "deep awareness of the history of great French and
British marques," the US was ultimately chosen because of certain similarities between
the Japanese and American markets, something our informant called an "exquisite
irony".49 Some of the common characteristics he mentioned: the two markets' passion for
novelty. in contrast to the seriousness of German customers and their focus on tradition
and refinement; the fast-paced agility of the Japanese and American markets, as
compared to the more slow-moving European markets, the French and British in
particular; also, the enthusiasm they share when it comes to technological innovations.
Nissan's decision to set up a design studio outside of Japan and to tap a non-Japanese to
found it and to run it, can be seen as a form of division of labor. This was understood by
Mr. Ishihara from the very beginning, as he explained his reasoning to our informant:
Look, there s a lot of things you guys over in the West are that we aren 't,
and we don 't want to trade who we are to he able to do what you can do.
49
The choice is even more surprising when one considers that in the late 1970s, very few interesting
automobile designs were coming from the US, where the Big Three had been struggling and spending
most of their resources to meet ever tightening safety and emissions regulations.
163
So let 's hold hands. You do what you do, we 'Ildo what we do, and let 's
figure out a way to do it together. (As recounted by Jerry Hirshberg.)
Jerry Hirshberg also understood the situation on these terms; he recalls Mr. Ishihara
saying to him: "We want to change the flavor of our soup; we're going to stay out of your
kitchen." And in retrospect, he feels that they have done just that for the last seventeen
years. Although, until a few months before our visit, the presidents of NDI had come
from Nissan in Japan, they were all told to "stay away from our process, even if it looks
weird." Jerry Hirshberg recounts the example when, in the middle of a workday, he
decided to take all the NDI personnel to see a movie. It was during the first Altima
project, and the design team was way behind; they "weren't hitting it." Both the Japanese
and the Americans were worried and the tension at NDI was high. Although the Japanese
president initially demurred, citing the status of the project and the timing of the outing,
the director ultimately prevailed on him to join the group.
Our informant summed up the decision by Nissan's president to create NDI thus: "He
(Mr. Ishihara) wanted an 'Americano', who would be abrasive in the culture, but in a way
that they could accommodate." He found him in Jerry Hirshberg. The exposure that NDI
and Hirshberg have received over the years, and the tight association of one with the
other both in automotive and industrial design circles, appear at odds with the anonymity
favored by most Japanese executives; yet that may have what Ishihara intended all along.
Hirshberg is invariably mentioned or quoted in virtually any news or magazine article
having to do with new Nissan models in the US. and even in articles that only deal with
competitors' products. 5 0 In addition, as mentioned previously, he has recently written a
book about his previous experience at General Motors, how he was hired by Nissan to
found NDI, and his management philosophy and practices there. These are described in
the next section.
50 A
typical example is an article in AutoWeek. dated August 24. 1998 (Autoweek. 1998). The article
mentions a number of companies. including Ford, Audi. BMW. and Cadillac, as well as Nissan. Whereas
mention is made of "sources" at some of the other companies, Hirshberg himself is quoted in the case of
Nissan. Another example is an article in Automotive News about the new Altima and its pricing and
positioning relative to Toyota's Camry (Rechtin, Crate, & Jewett. 1997). In that the article, Hirshberg is
mentioned four times, and Nissan's manager for the sedan line only once.
164
NDI as a Creative Entity
Before Nissan hired him, Jerry Hirshberg (JH) had been working at General Motors, as
chief designer for the Buick division. He describes his experience during his last few
years at GM as one of deep frustration and of "feeling more and more suffocated as the
70s droned on." As he sees it, GM, like many corporations, had many "senders" when it
came to creative work, but it was lacking in "receivers". "After a lifetime of having dealt
with just throwing ideas against the stone wall of a bureaucracy," he was determined to
organize NDI around the process of creating ideas as the end product. He would tell
potential employees "Idea-making is what we care about the most."
Initially, JH did not know specifically how he would go about creating that new entity; he
had an image in mind though, and the image was that of'"a sandbox and recess." He
knew that NDI would be built around the creative process, but he also saw limitations in
the ideas and prescriptions proposed by the numerous business books that deal with
creativity, which he calls "clever ways of manipulating, forcing, somehow opening a way
for creativity" within an organization. He resolved that his approach would be different.
At NDI, the human thought process, the creative process, would be the centerpiece of the
organization. Paradoxically, JH proudly proclaims that he was "smart enough not to try to
define what the creative process was, or [what] creativity was."
The result, according to JH, is an organization that operates differently, and "not just
differently than a Japanese company." "It's pretty odd amongst American companies
too." Several management researchers appear to agree, as NDI has been mentioned in a
number of studies, (Leonard & Rayport, 1997) and (Leonard & Straus, 1997) being two
recent examples.
Creativity: a marriage between the rational and the intuitive
Although he is careful not to give a closed definition of creativity, Jerry Hirshberg has a
very clear and strong notion of where creativity comes from. To him. "creativity resides
in the intersection of the rational and the intuitive" modes of thinking, "not in the
intuitive alone or necessarily the rational." The major impediments to creativity in most
organizations are thus two: one is the low status accorded to intuitive thinking in our
165
culture; the second is the tendency to divide and compartmentalize the different modes of
thinking.
JH believes that intuitive thinking is the most natural way of thinking, and it is the mode
of thinking on which we rely most often in our daily life. Unfortunately, these important
tools are "educated out" of us and, in the typical work setting where analytical thinking is
prized above all else. they are frowned upon as being "everything from 'sloppythink'
'feminithink'."
to
To JH, that is a mistake:
We do some of our best thinking when we sit with our wives and say,
Where should we go on vacation? Should we have a dog, or should we
have a baby? Then we think well. We come to work and we make charts.
We misapply scientific method to things that don 't have well-defined
bounds; but we have tools available, but they're not allowed.
NDI then is about "embracing [...] intuitive work." But as with creativity, JH is careful
not to give a simple, closed definition of intuition and intuitive thinking. Again, he relies
instead on some of its characteristics:
Whatever it is, we know what it ' not. We know that it 's not linear. We
know that it's not deductive.We know that it makesleaps that it shouldn't
have had the right to do. But so often, it s right.
The second roadblock to creativity, namely, our tendency to divide and
compartmentalize, is also a cultural phenomenon according to JH. It is ingrained in our
educational system, and it carries over into organizations where separate "creative
departments" are set up and tasked with developing creative and innovative ideas.
I think theproblem is that we tend to divideand subdividein this country
and in this culture.And notjust here; in the East as well. A lot of Western
educationis based on that. For me, all the great discoveries--allthe great
art, all the great science--is literally the flashpoint where the intuitive and
the rational meet, and it s very uncomfortable. So this is not an easy route.
Anybody who tries to neatly compartmentalize intuitive over here and
rational over here is going to run against the same wall, because that 's
going to make for 'creativedepartments whichare a blindalley as /ar as
I'm concerned
JH has therefore "developed strategies that attack that division head-on."
166
This kind of organizational (and often spatial) segregation of creativity into specialized
departments is only one type of mistake that organizations commit. Another is the
temporal segregation, when they assume that creative thinking is something that can be
summoned when needed, that is, when the organization is facing a challenge and it needs
to "break the mold." As JH put it, "The muscles aren't there; it's going to take a long
time." For that reason, NDI is organized so it is constantly immersed in the language, and
thus the way of thinking, of the creative organization:
The language of design, which is the language of intuitive and interpretive
thinking, along with rational thinking, has been introduced into the larger
methodology of the organization.
"Creative abrasion"
Ironically, at the same time that JH sought to eliminate the cultural divisions one
typically finds within organizations, between those groups that think and work intuitively
and those that operate analytically, he also became aware that cultural differences were a
rich source of innovation. This became obvious during NDI's interaction, on various
projects, with personnel from Nissan in Japan.
Something magical is happening in the struggle to communicate and
negotiatewith the truly alien culture.that is, betweenAmericansand
Japanese.It wasn't easy. And I realizedat somepoint that the struggle
wasyielding all kinds of innovation,if theprocess washandled in a
certain way.
This seeming contradiction disappears when one understands that it is the process
through which barriers are removed and conflicts resolved that is the source of
innovation. JH refers to this process as "creative abrasion."
But how are those ideas and concepts operationalized within NDI? Some of JH's
management strategies and methods are the subject of the next section.
Managementstrategiesand leadershipat NDI
Some of the strategies described by our informant are clearly intended to remove the
barriers that naturally develop between groups within organizations, while others are
clearly meant to encourage and support intuitive thinking. Others, however, do not fit
167
neatly within that classification scheme, as they appear to support both of JH's goals.
Some of the strategies were inspired by the arts, and not only the plastic arts, which have
a natural connection with industrial design, but by the performing arts as well. On several
occasions, JH reminded us of the emergent and evolving nature of these strategies. These
strategies are presented here, and, where applicable, the different type of leadership that
they call for is discussed.
"Stepping back from the canvas"
One example of this strategy, the NDI movie outing, has already been described earlier,
in the context of the Japanese president' role at NDI. This strategy was inspired by
something that one of JH's art school teachers used to ask his students to do. When
painting, and after working very closely to the canvas over an extended period of time,
one develops a distorted view of his or her work. By occasionally stepping back from the
canvas one can see the result of one's work from a more realistic perspective. From time
to time, and especially when a project is facing difficulties. NDI shuts down and
everybody goes on an outing, to the movies in particular.
This strategy goes against the traditional approach where. when facing difficulties,
managers will hold a meeting. remind everyone how important the project is, and call for
overtime. Or, as JH put it, "Noses to the grindstone; buckle down, get to work..." lie
explained:
What most leaders don 't realize is that work in and of itself is intrinsically
an inward-turningprocess.It pulls you into it. It 'smagnetic.I s a stupid
waste of time, and harmful, when leaders and managers push-you know.
Get down to work! The trick is to strategically-and you have to know the
right time-pull peopleaway.fromwork.
Usually, people use the weekend or take a vacation to "interrupt the rhythm". to change
their thoughts and to be able to come back to their work with a fresh eye. At NDI, this is
done during work, when JH believes it is most effective. And it does not always involve
an outing; it could involve one designer taking a break from his or her project and getting
involved in another designer's project.
168
"Cheating"
"Cheating" is a practice intended to break down barriers between people, a "deliberate
attempt to encourage interference, intrusion, sticking your nose in other people's
business." When young employees are hired into NDI, they are told that they are
expected to "cheat" from one another. They are expected to rifle through each others'
papers and drawings. They are also told that their colleagues will be rifling through their
work. They are invited to do it when the other person is away at lunch, if that makes them
feel more comfortable. This is another idea that JH picked up from the arts; in this case,
he credits music where many pieces were inspired by other composers' works, or, as he
puts it, "the best ideas are stolen ones." He admits that he is using "stolen" in a special
sense, that "inspiration" would be another good way of describing the process too. He
gives another example from music, specifically jazz, where one musician in the ensemble
would do a riff on something that another just played, "... like a sax hearing a drummer
and doing something with it." Of course, the person doing the stealing is expected to give
credit and to give the idea his or her "own spin."
JH credits this practice for the idea for the Nissan Pulsar, the first mass-produced car to
offer the customer modularity, whereby the customer could convert the car from a minitruck to a notchback to a hatchback by swapping body modules. It should be pointed out
that this is an extreme case of the practice however, as it involved one designer
misinterpreting another's work, not merely reinterpreting it. The idea for a modular car
came to JH as he was looking through some of the drawings that another of the NDI
designers had been working on. The designer had done a sketch in which he had
delineated various body panels with thick black lines. While rifling through those
sketches, JH misinterpreted the thick lines to mean that the outlined panels could be
removed and replaced with other ones, like a three-dimensional puzzle. Mistakenly
thinking that the designer had been working on a modular vehicle, JH did some sketches
of his own, and the Pulsar was born.
The practice of allowing and even encouraging this type of "cheating" within an
organization that thrives on creativity places some heavy demands on the leadership of
that organization, in particular when it comes to fairly apportioning the credit for a given
idea or creation. From the interview, it seems that NDI does not rely on some intricate
169
accounting mechanism to solve this problem; rather, rather, JH has created the kind of
closely-knit culture and work environment where such concerns are effectively
:,idestepped. He explains: "We won't put up with somebody saying, I did that first, that's
mine." At NDI at least, this approach seems to work, as he goes on to say: "It's amazing
how that becomes a non-issue."
"Blurring the boundaries"
This strategy is another "deliberate attempt to encourage interference, intrusion, and
sticking your nose in other people's business." The idea here is to eliminate the generally
accepted division of labor, whereby designers do their thing and engineers do theirs,
without either telling the others how to do their job. When engineers are hired at NDI,
they are told up front "We [the designers] are going to step on your toes, and we want
you to step on ours." To JH. in any given product. the contributions of the designers and
those of the engineers "are so overlapped that we don't even bother to pull them apart
much."
JH gives the example of the Quest minivan project as a case where that strategy paid off
handsomely. 5 ' Initially, the designers were not very enthused about designing a minivan;
to them, doing a minivan seemed "kind of boring." They responded by trying to make a
minivan that would look swoopy. The early sketches depicted one-box designs that JH
referred to a "rocket ships". (A one-box, or one-volume, minivan design is basically a
design where the front section of the vehicle, where the hood and engine compartment
would normally reside, is not clearly defined or delineated from the rest of the vehicle, as
it would be in the more common two-box design. The result is a vehicle where the A-
pillars practically form a straight line set at a very shallow angle with the horizontal,
resulting in a very large windshield, a windshield base that is very far forward from the
passengers, and rather poor visibility for the driver and front passenger. The first GM
minivan, the Pontiac Transport of the late 1980s was a one-box design, and it was
criticized for being overstyled, for lacking functionality, and for its A-pillars that
obstructed the driver's view. An example of such a design, perhaps a more successful
5s
The Quest nlinivan was a joint development between Nissan and Ford. t was marketed in the US by
Nissan as the Nissan Quest, and by Ford as the Ford Windstar. The two models differed only in detail,
and NDI was responsible for the design of both.
170
one, is the current Renault Twingo.) Upon seeing these sketches, one of the engineers
suggested they try to see what it would feel like to sit inside such a vehicle. So designers
and engineers got together and built a plywood skeleton of the one-box van design,
complete with wheels and seats. They then sat in the skeleton prototype, and it became
clear to everybody that it was not a good design:
... we sat down in it, and that was the death of the one-box car. We said,
'This is ridiculous. Look where the A-pillars are!'
JH credits that fortunate outcome to having engineers onboard at NDI, ready to "interfere
with the design process, and the designers not saying, 'What the hell do you have to do
with this? Go do engineering work!"'
"Hiring in divergent pairs." The Quest story above is an example of"creative abrasion",
in the sense that the idea to move away from a one-box design was the result of the
engineers, or one engineer, challenging the thinking of the designers and the direction in
which they were headed. But for "creative abrasion" to work, the organization needs to
bring in people who are different in the first place. This is the role of this hiring strategy.
JH describes it this way:
We always hire in twos, whenever we can. And whenever we do, we will
deliberatelyhire about as oppositionalas we can think Whoeverthe hell
it would make that person uncomfortableto be in a phone booth withfor
any length of time: different schools, different cultures, different countries,
different philosophies, different methodologies.
Obviously, the benefits of hiring in divergent pairs would be short-lived if the employees'
differences and unique backgrounds are watered down when they start interacting with
one another and with the culture and the way of thinking of the organization. JH is aware
of the need to prevent such dilution from happening:
... we spend a professional lifetime encouraging these folks to continue
working and thinking the way they came in; and honoring that, and
celebrating that.
And:
171
My hope is that other people are operatingin their own individualways in
this company.
JH did not have any simple answer or strategy as to how NDI does that, but his
description of how NDI operates makes it clear that conformism is not what NDI expects
from its employees. People are encouraged to ask questions, no matter how dumb they
might seem ("we encourage people to make fools of themselves, not by being idiots, but
by asking dumb questions.")
The way product critiques are conducted at NDI is a good case in point. In reaction to the
way critiques are conducted at Detroit, NDI established some unwritten rules from the
very beginning: when a product is about to be shown internally, the event is announced to
everyone over the PA system; anybody who is interested is welcome, on the condition
that they openly say whatever comes to their mind. "Anybody" includes designers from
the other design studios, from the color studio, the design context laboratory (described
later), the model shop, engineers, administrative personnel. and. occasionally, even
visitors. By contrast, the practice at other car companies is to invite only designers to
critique a new design. This point was made clear during the Chrysler visit, in the
interview with Chrysler's Director of Exterior Design (see Chapter IV). who stressed that
only "people from the Design Office, the designers and the modelers [... ] people who are
sensitive to design, trained in design, were allowed to vote on a design.' In fact,
according to JH, Chrysler's would be a relatively progressive design office, as his
description of the traditional Detroit approach as follows:
In the more progressive studios occasionally. some engineers and
modelers were invited, but they were not expected to say anything,
because this was a design critique. It was only from the studio, and only
those people working on that project within the studio, with the exception
of the executives, who also came to the viewing patio and viewed these.
The rules were very clear: rank spoke;you spoke if you were asked a
question;andyou only said somethingnegativeif you had a better
answer.
To NDI's way of thinking, that practice seemed "stultifying" and "horrible". NDI's
approach, on the other hand, does not come easily to designers, and especially to heads of
design studios. Traditionally, automobile design chiefs have tended to be individuals with
172
strong and sometimes colorful personalities. Generally, they wielded a lot of power,
acting almost autocratically when it came to design decisions. The archetypal dictatorial
design personality was Harley Earl, General Motors first head of styling; more recent
examples would probably include Chuck Jordan, GM's former design chief.5 2'53
As an example of the different kind of atmosphere that he has tried to foster at NDI, and
of his different style of leadership, he described a meeting during which an unexpected
idea was proposed, and he felt that it should be adopted. One of his subordinates
interjected that that would be "wrong", that it would be better not to do anything about it
right then, but rather to think of it as one question within the discussion they had had up
to that point and to sleep on it.
... we 're in a culture where you can say that to the boss. [...] it s very
important in an idea-making culture to accommodate leadership coming
from places other than the boss. This is corny, but good leadership in this
situation is goodfollowing.
The above example highlights a very important aspect of the environment at NDI,
namely, that it supports open-endedness. Autocratic leadership or "strong" leadership in
the traditional style is, in a sense, at odds with open-ended thinking. It calls for decisions
to be made quickly and without equivocation. The leader is expected to know the answer.
The following section describes other practices that support and even encourage openendedness, a key feature of interpretive management.
Strategiesand practicesthat encourageopen-endedthinking
Some of the strategies already described in the previous section have elements of openendedness. For example, "creative abrasion" and "hiring in divergent pairs" are strategies
the outcome of which is not known a priori. Although it may be obvious that the
interactions between an engineer and a dcsigner are likely to yield a more workable
product, it is not clear what the details of this outcome are likely to be. The situation is
even less clear when the difference between the different people who are being brought
52See (Bayley, 1990).
53
Wayne Cherry, GM's new head of design, is the antithesis. He is described as being modest and having a
"low-key" demeanor. See (Child, 1996a, b).
173
together relates to their cultural and educational backgrounds. At the time of the our visit
to NDI, their "most recent hires [were] from Germany, India, England, Canada, and
Iran." Clearly, these people were hired on the expectation that something interesting may
come out of bringing them together, but that something is not obvious and it may not
even be discernible before the fact. Hiring an Iranian and an Indian do not mean that
Nissan intends to bring out a product with Middle-Eastern or Indian styling cues.
Following are some other strategies that favor open-endedness.
"Porous planning." The idea behind "porous planning" is the realization that, in a creative
entity like NDI, one is typically making choices between options that are still ill-defined,
and that it is often difficult to separate the decision point from the creative moment. Real
life does not always conform to the generally accepted analytical model of design, which
consists of a series of "generate/test" cycles, where in the first phase a number of
solutions are generated and, in a later phase, these are evaluated against a set of
performance criteria and the solution with the highest score selected.
It is not clear from the interview notes how "porous planning" is actually implemented.
From the example given by JH, it appears that "porous planning" is more an attitude, a
willingness to change directions at a point in the project when the relevant design
decisions have been fixed, rather than a specific planning methodology that allows for the
fluidity of creating novel designs. The following quote refers to the previously mentioned
Pulsar NX project, after JH came up, accidentally, with the idea for a modular vehicle:
This was mid-product, so it had to do with the unexpected being injected
into a well-definedplan. Nobodyde.finesa plan better than the Japanese.
But fortunately,we had a wonderful "Shukan," project director...product
planning engineer, who went with theflow, who said, Wow, eureka, let 's
do this. And we literally shified directions midstream.
"Creative questions before creative answers." In JH's view, intuitive thinking places a
premium on asking questions. At NDI therefore, people are "held up from answers for as
long as possible." At the same time, he recognizes that this is exactly the opposite of what
people need in order to move forward with a project and to meet deadlines:
174
One of the things about intuitive thinking [...] the stress here is on
questioning.We holdpeople backfrom answes for as long as possible,
which is the exact opposite of what everybody needs to do to finish that.
We like to tilt the plane, and leave it tilted, and live with the ambivalence
just as long as you can.
The focus on open-endedness however goes beyond delaying answers until more
questions have been asked; it directly affects the nature of the questions being asked. The
clearest examples relate to how NDI conducts its customer research. (This topic will be
covered in greater detail in a separate section on the Design Context Laboratory.) Very
often, consumers, either Nissan customers or people who already own competitors'
products, are invited to come to NDI and bring their cars along. These sessions however
are unlike the typical fbcus groups where a list of specific and detailed questions is
prepared a priori, and where descriptive words are chosen carefully to gauge the
participants' response to different design proposals. (See Chapter V above; specifically,
the interview with Chrysler's Director of Design on how particular descriptive terms are
suggested to focus group participants in order to classify different design proposals.)54 At
NDI, there are no prepared questions. JH describes these sessions as follows:
What we do is, we call people and say, 'Listen, we understand you're an
owner. We 'd like to meet with you. We '1lgive you a small amount of
money--fifty bucks or something--and we want you to come with your car,
and leave it as it is:you come dressedas you are, and we 'lljust live
togetherfor a couple of hours. ' We have no questionsthought up
beforehand, and it 's not important to us either--which is another thing to
break the rules--thatwe show results:that, in other words, "This better be
good" or "We need to get somethingtangibleout of this." We never know
what the hell we 're going to get, and sometimes it 's nothing. Then you
know you 're on the right target. So we get together sometimes, and we '11
begin observing,andjust kind of dancingaroundissues, maybe talking
about the car, and maybe not.
What is remarkable in the above description is how the process itself seems to be more
valued than its outcome. Whether anything tangible comes out of it, the exercise is seen
54
The Minivan team at Chrysler did conduct the kind of research described here. The team members talked
about looking at what minivan drivers carried around with them and so on. The distinction i'm referring
to here relates to the methods used by Chrysler's Design Office (please refer to Chapter IV.)
175
as valuable because it would have allowed the designers to get to know their customers
better, even if that added knowledge couldn't, at the time, be articulated or translated into
specific product characteristics. To use an analogy from interpretive thinking, it is as if
the designers are trying to discover, or to learn more about, the "horizon" or the
worldview of their customers. There are cases, however, where specific ideas do come
out of such sessions, as in the following example that was related by JH:
Anotherwoman had threepillows of differentsizes in the car ...One was
for driving, one was because she loves to take a nap at noon, and one was
because she likes to visit somebody, and the chair is uncomfortable. There
were three difjerentpillows, and we started thinkingabout thepossibility
of modular removable pillows, and things you could take withyou. The
opportunitiesfor innovationarejust so much richerwhenyou 'releaving
the discussion open.for the unexpected,fbr the unexplainable,Jbr that
which... In the nature of creativity: the most uncomfortable part of it is
precisely thefact that you didn 't think to even ask the question that led to
it in the first place. Ifyou did, you 're there. To our way cafthinking, the
creative question is the key.
Obviously, NDI researchers do not go into these sessions with totally blank minds. They
are always already involved with a particular project or problem, but they are careful not
to have already formulated answers or categories in mind. JH describes it as "going into
the field unprepared; deliberately but professionally unprepared." The crucial point is not
to go in with "'a question that presumes an answer."
Work on remaining amateurs; drinking from diverse wells. This strategy is an excellent
example of how JH is careful not to let NDI sink into the hubris that has often
characterized the successful design studio and the autocratic studio head. JH sees this as
especially important when designers are developing the next generation of a successful
product or, more generally a new product in an area in which they have had success. For
example, one of NDI's first successes was the Nissan Pathfinder, which played an
important role in creating the current market for sport-utility vehicles (SUVs); at the time
of the interview, several NDI designers were working on a Nissan's upcoming entry into
the small sport utility vehicle segment (a relatively new segment now occupied by
Toyota's RAV4 and Honda's CR-V.) In explaining the reasons for sending some of these
176
designers on a multiple city tour to conduct focus groups, JH stressed the importance of
thinking like an amateur:
...we always try and strip ourselves bare of any presumptions qf expertise,
especially when we 've been successful in the past. So we work very hard
at remaining amateurs here; that s one of the reasons we do diverse
products--keep ourselves again lipped, uncertain. It 's healthier,
It is interesting to note that Chrysler's Director of Design was very comfortable with the
fact that his designers only work on automobiles, whereas JH sees NDl's non-automotive
projects as playing an importance role in maintaining his designers' minds stimulated and
open:
I talk about the valueof a companymakingcertain that its employeeshave
opportunitiesto work onprojects outside the realm of their expertise,from
time to time. It doesn 't have to he continually, but it must be from time to
time. So the connection between the projects, also, is the health-giving and
is the stimulation that comesfrom working on a minivan here, and then on
anotherpart of the designers' deskyou '11see a golf club or, currently,a
laptop computerfor [CompanyX]: or a sportscar and a 105-footyacht,
or a minivanand a chairfor kids. We havefioundremarkableand
unexpected (always) crossovers, but we don 'tfobrcethe issue. So it's kind
of acknowledgingwhat everybodyknows and says aboutcreative
breakthroughs: they inevitably seem to happen from somewhere else, and
it 's usually unexpected:.from a neighboring field We 've brought the
neighboringfields home.
In this section, the focus has been on certain strategies at NDI that encourage openendedness, a key aspect of interpretive thinking. The next section discusses other
practices that are more directly interpretive.
Strategies that encourage interpretation or interpretive thinking
Blurring the boundaries, while keeping people distinct. A hallmark of hermeneutic
interpretive thinking is the notion of the 'fusion of horizons' of the interlocutors involved
in an exchange, as they 'come into understanding'.
Each of the interlocutors comes to the
encounter with his prejudgments and background knowledge and beliefs, that is, his
'horizon', and at the end of the successful encounter, each will have had his 'horizon'
altered to some extent as understanding is achieved. Understanding does not mean that
177
one of the interlocutor ended up seeing the situation as the other saw it at the outset;
understanding is not achieved by critically eliminating what is foreign, nor by simply
uncritically accepting it, but by translating it into one's own terms.
The strategies described previously, "blurring the boundaries" and "hiring in divergent
pairs", coupled with the efforts at NDI to prevent its people from falling into conformity
and to retain their preferences and values, are key to the interpretive flavor of the NDI
environment.
Personalysis: providing interlocutors with information about each other's 'horizon'.
Personalysis is a tool used to characterize people's personality along a number of
different dimensions, such as action-orientation, thinking orientation, people orientation,
the importance of history and continuity. The system used at NDI is very visual, relying
on colors, red, yellow, green, and blue, respectively, to code for the different orientations.
The system is somewhat more complicated, as it attempts to quantify these orientations as
the person faces different situations. The result is a graphical map of a person's
personality. While admitting to the shallowness of such methods, our informant presented
the system as an important tool used at NDI to facilitate communication and interaction
between people. Each person has on his or her desk a three-by-five card with that
person's 'colors' on it, for everyone to see. This encourages people at NDI to take into
account their interlocutors' ways of thinking and values when they interact with them,
something that is very important in an environment where people as diverse as color
designers and mechanical engineers are brought together. Our informant put it in
Heideggerian sounding terms: "... what works is that we're thinking about somebody
else's way of being in the world." The other hermeneutic aspect of the Personalysis cards
is that they are another way to encourage people to keep their particular idiosyncrasies
and ways of thinking: "It's another means of retaining each individual's color
characteristics, core characteristics. and kind of honoring them."
The designer as '"method actor", and "embracing the dragon." Hermeneutic interpretive
thinking is also evident in the methods devised by NDI to understand their potential
customers. JH describes it as becoming method actors, in the sense that the designer
attempts to experience the life of the customer:
178
We become method actors, rather than market-savvy. So when we are
asked to design a truck, we become truckers. We do that by, (a) we get a
bunch of trucks in that are considered interesting; we 'll usually get
together and say, " Which ones do we want to drive? " We '71get them in
here. [...] When we 're asked to design a van, we 'll all become van drivers,
and we '11let ourJAimilydrive them. We'll use it, and we '11start griping
aboutit, and whiningaboutit, andjoining the opposition.We call that
"embracingthe dragon." But really getting into the process, and suffering
the product along with everybody else. And admiring those parts of some
of our competitorsthat are done damn well.
Clearly, "method acting" and "embracing the dragon" allow the designer, one of the
interlocutors in the exchange between designer and customer, to get a glimpse of the
other's 'horizon', to experience the other's world, as an important step towards
understanding his or her needs.
Analogic and metaphorical thinking. Method acting and interpretive thinking are not
limited to the interface between the design organization and the product's customers or
potential customers. It is also used internally as the design team tries to imagine what
their product might look like. The following quote is a good example:
... what 's a van? t 's a mobile, flexible, interior large volume on four
wheels. That 's what we defined it as. [...] We regard [what we came up
with] as a handsome toolbox. It should be like--what 's the tool company
that does nice design? Black and Decker. There is a company I think that
has done some handsome design over the years. We said, If Black and
Decker did a van, that's what we 'd want to do. That's how we kind of
defined it. Or if Porsche did a piece of hand luggage,that 's this.
The quotation reveals complex. multi-level analogies. At the first level, having rejected
the idea of a minivan as a swoopy "rocketship", the designers start thinking of a minivan
as a toolbox. Then, thinking about the tools of one particular company, the designs of
which they like, they try to imagine what a minivan designed by that company might look
like. In a way, this is similar to method acting. where one company is trying to
experience the life of another, the character, in this case Black & Decker. The second
example, involving a piece of luggage and another company whose designs our informant
apparently admires, namely, Porsche (or Porsche Design). follows the same idea.
179
Having looked at the organizing principles and the practices of NDI in general, and the
underlying thinking of its president, we next turn to a specific design project and to the
people who were responsible for it on a day-to-day basis.
The Work of a Design Studio: The Altima Project
The interviews at NDI were wide ranging and covered a number of general topics
including the history of the organization and the management and design philosophies of
its president and co-founder (see previous section). Still, the original focus of the visit
was the development of the (then soon to be released) Nissan Altima. To that end, the
chief of the studio responsible for the Altima design, henceforth referred to as D1, and his
assistant during that project, referred to as D2. were interviewed. At the time of the
interviews, D2 had herself been promoted to chief designer, managing another of the NDI
studios.
The organization of a design studio
At NDI, a design studio encompasses all the different technical disciplines necessary for
automotive design: design, modeling, and engineering. A studio has an engineer, a few
modelers, and a group of designers, typically three, including the chief designer who also
manages the studio. This arrangement was described as unusual, as many companies do
not have a designer acting as a manager too. The fact that the chief designer/manager of
the design group is actively involved in the design work was described as significant:
Yeah,the chiefdesigner-managerof the group is on the [drafting]boards
too; he 's participating. That s what we wanted to do here; we determined
thatfrom the beginning, that there will be no 'hall-walkers'here.
[Laughter.]
The informant, Dl, went on to compare design to a language and. specifically. to a
foreign language. Without regular practice, one's ability to speak the language will
decrease over time and ultimately disappear. Similarly, a manager who does not spend
any time designing will "lose touch with the working of design, the actual work, the
building of a body of knowledge about form and design..." This is especially critical in
180
an environment like NDI, because drawing and sketching are the language used regularly
to communicate and exchange ideas with other people.
The respondents contrasted the organization of their design studio with that of the Nissan
design studios in Japan. Compared to NDI, the design operation in Japan is much more
structured. The different disciplines are not located under one roof, in one large area.
Rather, the designers are housed in a small room off to the side of the studio area proper,
and the engineering staff are in a separate engineering building, necessitating frequent
trips to the studio to exchange information and data. The studio area itself was described
as being the domain of the modelers. Another point of difference between NDI studios
and their Japanese counterparts is the environment; Japanese studios were described as
being "quiet as a church." By contrast, the atmosphere at NDI is playful, with music
playing, people telling stories and laughing as they work, and even throwing balls of
modeling clay around.
A designerand his creation
The respondents identified two major phases in the gestation of the new Altima. The
initial brief called for a design that would share many components with the larger Nissan
Maxima sedan. That phase was carried all the way to a full-size clay model that was
taken to Japan for evaluation. That project was ultimately canceled because the
manufacturing organization, NMA, concluded that the resulting design would have been
too expensive to build, despite the economies achieved through the sharing of parts with
the Maxima. In addition, the overall dimensions of the proposed model were too close to
those of the Maxima, and Nissan's marketing organization, NMA, was concerned about a
bigger Altima cannibalizing sales from the Maxima, a model that had been very
successful for them.
This led to the second phase, in which a replacement was sought, which was closer to the
original, first generation Altima. Dl's studio at NDI was contractually required to deliver
three quarter-scale clay models, representing three different design directions for the
second generation Altima. D2 explained that, as sometimes happens to designers, neither
181
she nor the third studio designer could come up with a design that "flourished", so Dl
ended up doing all three scale models. 5 5
In describing the scale models and the process through which one of them was selected,
the informers focused on two of the three. The first was described by D as having a "real
kind of art-deco thing going" and as being "really, really different", going beyond the
Infinity J30 along the retro dimension. Although DI referred to it as his favorite design,
and D2 characterized it as being a "big favorite" and "a lot of people's favorite", they
also referred to the design as "bizarre" and I)1admitted that it was so different that
"people were kind of afraid of it." The second scale model, the design that was chosen to
be developed into a full-scale model, was described by DI as "fit[ting]everybody's
image
of what the next Altima should be." The final selection process, as described by D 1.
seems to have been an unusually straightforward one, at least as far as the executives
were concerned. The final show was held in the courtyard at NDI, with all the quarterscale models set up on display. Executives from NMA and NMC came from Los Angeles
to San Diego for the event.56 DI presented the different designs and discussed the merits
of each of them. That was followed by a vote that was asily carried by the second
design. Not only did that design get the most votes, but there was also a consensus among
the executives from NMA and NMC. as well as Jerry Hirshberg. that it was the design
that should be pursued further. D2 described it as an easy decision, because of the
consistency of preference shown towards that model, initially among NDI employees
who were polled during earlier shows, and later among the Nissan executives and the
NDI director.
But this easy decision belies what could be described as a certain amount of anguish on
the part of the studio chief. Although he realized that the 'art-deco" design was too risky'.
that it had little chance of being selected, and that the second design was the less risky
way to go. he was still attached to it and he was still hopeful that it would be selected.
..s5 sometimes designers are in sync and they totally run with the project and other times. you just don't
have it, you don't have an idea."
5'
The courtyard in question is one of the more unusual features at NDI. often singled out when the
architectural layout of the NDI building is discussed. It's an internal court that makes it possible to
display models and vehicles in the open, under natural light. while keeping them safe from any prying
eyes.
182
That had been the case from the beginning, before the final show in the courtyard. That
model started with something he had observed his wife doing when she got in a car.
Being barely five feet tall, she has to move the seat all the way up and forward so she can
see the road and reach (barely) the pedals. Once, after she got in, she tried to pull the
visor down to check her makeup, and the visor hit her head, which upset her and made
her give her husband an earful about how cars are designed. Specifically, she complained
that, by the time she gets the seat up, not only can she barely reach the pedals, but the
steering wheel is too close to her chest and the sun visor too close to her head. D
understood that that was the direct result of the never-ending trend in automotive design
to make cars sleeker and windshields "taster". that is, more horizontal. Dl's reaction was
to start thinking about the implications of making the windshield more vertical. The
resulting design was "like an older statement, from a different era." DI developed a
model with a more vertical windshield and a teardrop-shaped side glass, resulting in, as
he put it. "a whole new, different dynamic in the vehicle." But from that very beginning,
people's reaction to DIl's favorite model was negative:
At first I remembertaking that modeloutside and people i'ere lookingat
have you done?
it [and saying/. Whatare you thinking?PWhat
The scale model shown to the executives was not that first model but a second iteration in
which D1 had "tightened up" the design a little bit and made it "a little more mainstream
looking". while still keeping "that gesture". Despite the easy win of the second proposal
in the final NDI show. DI was still hopeful that the situation would change and the artdeco proposal would be reconsidered for a full-size model. The NDI show was to be
followed by a similar presentation to a committee in Japan. However. DI 's hopes were
dashed very early in the meeting:
I was asked at that meeting by some of the designers there WhyI brought a
car that was so deliberately ugly. You know. So. I knelt that that one w'as
not going to fly: that was it. it wac.all over. Because. I always had hope
that maybe I'd get over there and they would sa, That :' it! Or. you know.
You got it! And then I would come back and I'd have some leverage to
possibly sell it.
183
And so this spelled the end of the art-deco design. which was both D l's and D2's
favorite. D1 was philosophical about it:
Sometimes, you know, you get an idea and the idea is so clear in your own
mind, but the vision isn't... everybody can't see i. And that's how it works.
Sometimes yotuhave something that everybody kind of gets on board with,
and sometimesyou 'rejust out therein lefifield somewhere;you're having
imn but, you knowt...
A committee car?
After the winning quarter-scale model was selected, the design studio team faced the task
of developing it into a full-scale model. D2 described the scale model as having "the
smell of the car" or "the essence of the car", meaning of the final design, with its "lean
surfaces and edges". Still, a significant amount of work still needed to be done. This
phase of the process was described by D2 as one of the elements that make NDI unique.
namely, that they "work so well as a team. without the ego dominating." Although the
three designers had not worked together previously-D2
got moved to DI 's studio after
the Maxima project was completed. and the third designer was practically straight out of
school-all
three attacked the full size phase together. D2 did the rear quarter of the car.
including the quarter panel. the sail panel. and the taillights. According to Dl. "that gave
it life" because at that point. the car had "movement". and "the ultimate gesture was back
into it. and it had the right stance." The third designer contributed some ideas to the rear
too, and to the bumper area. D2 also worked on the front of the car. at different times
moving between the front and the back. D2 went on to describe her experience working
as a team in glowing terms:
It was .! wonderil experience as a designer to he involved in the
teamwork working on a.full size. and the way [DI] was here----itwashis
original idea and he was so open to the participation of his staff-and how
it worked to integrate in the car'... It just kept moving ahead which is kind
of interesting.
Given that the final product was the work of three different designers. D2 was sensitive to
the fact that the Altima could be called a "committee car". a harsh descriptive in the
individualistic world of automotive design. T'hatnegative expression is usually used to
describe designs that have been ruined as a result of the interference by powerful
184
interests, such as a powerful marketing department or executive requesting changes to
specific design features. D2 described the situation with the Altima and the three
designers working in concert as a wholly different phenomenon:
It's terrible to say it was a committee car, hut in actuality, it reallyflowed,
asfar as ideas here... We were really, how would I say, meshing well.It
was like we allJelt the car, what it needed; we all had a desire to mature
the existing A Itima...
Clearly, the three designers had reached a point where they could all work together
seamlessly, completing and elaborating, but also extending and modifying each other's
ideas. The process they describe has the hallmarks of a hermeneutic exchange. When D
developed the scale model, it is not that he had a fully evolved design in mind that he
communicated incompletely or ambiguously, and it was not just a matter for the other
designers to uncover and understand what he tried to say or show. Otherwise, D2's
contribution could not have been so important as to "give life" to Dl's design, or to
change something as basic yet overarching as the "gesture" of the car, to use D1's words.
D2, and the third designer too, played an active role not merely in uncovering Dl's (preexisting) idea, but in forming and modifying that idea.
D2 explained that the three designers had a shared desire to "mature the existing Altima
to take a bit of the goop factor out of it. and to make it more expensive looking because
these buyers love value for dollar." A 'so, they all wanted to make the Altima more
"appealing from a masculine standpoint, [that is.] stance and posture", as the previous
Altima was seen as too feminine by many. These shared objectives may appear to go
against the hermeneutic notion of an eventual fusion of initially distinct and differing
horizons, since they seem to indicate that the three designers already had a shared
'horizon' to start with (at least as for as it impacted the design of the car.) However, these
shared notions were at a more abstract level than the operational level at which the
creative exchange described in the previous paragraph took place. The notions of"taking
some of the goop factor out" and of giving the car a different "posture" or "stance" are
perhaps less abstract to a designer than the concepts of "more expensive-looking" and
more masculine, but they do not carry with them one answer or one solution to the design
situation, but leave a lot of room for interpretation.
185
A family or a supermarket? What makes a Nissan?
The decision to choose the more conservative design proposal at the quarter-scale model
stage highlights the tensions that design organizations typically face when developing a
replacement for an existing product, especially one that has been successful. Although
some of the younger designers at NDI wanted to take the new Altima in a completely
new direction, to do something "really wild and very different," other designers, and the
decision makers at Nissan, were swayed by other considerations, all of which reflected
their concerns for the identity of the product. of the company, and how they are perceived
by the customer.
When the replacement project was started, the original Altima had been on the market for
only a few months. It was beginning to sell well, and it was becoming clear, even then.
that the unusual design was going to be a success. The design was unusual then because it
was unlike other vehicles on the road, and it was a total break with the boxy. angular
Stanza which it replaced ("a huge leap" in the words of our informants.) By comparison
to the Stanza which had been a poor seller, the Altima had rounded. curvy surfaces, with
hints of retro-styling elements, especially in the rear. In talking to those early Altima
customers, Nissan had tfound quite a high level of "commitment" to the car and its
unusual design, and several Nissan designers. as well as the Nissan marketing executives.
felt that they should not alienate these customers by going in a different direction with the
replacement model. D2 explained:
... we didn't wcantto alienate them hvi.ust going on a diferenl tack,.so w'e
definilely wtnted to honor whal they loved about the car, hut improve
wha't 'wasstoppingotherpeople from hiuing it...
And D 1:
,nd. wefell that because of that success.l...j making that car so dilfjren
firom the previous car, the Stanza. that now if it had gone in a new
wanted to
direction--and we didn't want o confi se the customer. 'weC
establisha look.forthe car that iwouldcarry somethingof the old car
through, would have some consistency. that the cu.,tomers could relate to.
!lroducl. it had developed its own
It wouldn 't he seen as another kind of
look and now we wanted to mine that a little bit. and use it. And of course.
VMC'and Sales and Marketinggroup and Researchgroup reallvi/JIt that
186
that was a wise way to go, and there was a lot ofpressure from them to do
that.
In describing that decision, our informants kept making the point that, as much as they
liked the more daring, art-deco proposal, they felt that going with the more conservative
proposal was the wise thing to do from a business standpoint, that it was the sensible,
rational decision, the "responsible" decision, which was the managerial point of view.
Again, Dl:
And then there was the management [at NDI], and they flit that no. it was
time to be a little more responsible. to look at the long range business
plans, to develop the image of the car ...
This tendency towards caution that Dl describes is ironic: the success of a daring,
unusual design often causes decision makers to become cautious when the time comes to
develop a replacement, and to choose a more conservative design, thus destroying the
original source of their success. This tyranny of the successful design was mentioned by
other informants in other case studies, as we shall see in the next chapter on Nissan's
Technical Center in Japan.
The Design Context Laboratory
The previous section focused on a particular automobile model and the design studio that
was charged with developing it. In the next two sections, we focus on two entities within
NDI whose work cuts across the projects normally undertaken by the three design
studios. These are the Design Context Laboratory, the subject of this section. and the
Color Studio in the next section. Both of these fall under the same manager. and they
both very consciously rely on highly interpretive approaches to perform their functions.
The history of the DCL
The DCL was established at a time when NDl's relationship to the parent company
started evolving. Early on, NDI was intentionally kept isolated from the Nissan planning
and marketing operations in Los Angeles. The idea behind that decision was to keep the
designers free from the pressures of Nissan North America (NNA). NDI communicated
directly and exclusively with Japan. In time however. as NNA restructured, and as NDI
187
was given more responsibility for the design of products. especially those aimed at the
US market, the need for NDI to interface with the other parts of the corporation in the US
increased. In the course of its interactions with the US parent company in Los Angeles,
one individual at NNA was "very receptive and supportive" of the design activity at NDI.
He was more forthcoming with information than other managers at NNA and Nissan
Motor Company (NMC). Later on, he was invited to join NDI, where a position was
created for him. One of the tasks with which he was charged was to set up a colors and
materials studio, and another was to create the entity that became the Design Context
Laboratory.
According to our informant, everyone who was involved in its creation probably had a
different idea of what function they wanted the DCL to perform. The objective was borne
of a nebulous idea" that some kind of a group was needed which would provide research
information to the designers in a form that was useable and useful to them. and which
would play the role of advocate for the work of the designers within the corporation by
speaking the language of business people while being "design sensitive",. that is. able to
speak the language of the designers.
In addition to its manager. the DCL consisted of two persons at the time of our visit: our
informant, who had been running the day-to-day activities of the DCL for five years, and
another individual who had been hired about one year prior to our visit. In view of the
unusual role they were intended to play within the organization. it is instructive to
mention the backgrounds of these individuals. Our informant started out studying
literature and philosophy: he later switched to art history. which he studied in France.
After that, he worked in journalism and studied publishing. Before he was hired by NDI.
he was the editor of a well-known magazine on industrial design, targeted at
professionals in that field. The other member of the DCI, is an applied linguist by
training. although he is often referred to by some NDI executives as an anthropologist.
due to the kind of research he conducts and the methods he uses.
The DCL and internalintegration
Our informant explained one of the main roles of the DCL. by contrasting
the mindsets or
worldviews of the business person and of the designer. Even though designers understand
188
that they are working within a competitive business environment, they will generally try
to explain their work and advocate for what they have done in "designers' terms", that is,
in a language that is based on art. Their daily language is "the language of form, the
language of creation, the language of line," and the language of color. It would be very
difficult for a designer who has been so close to his or her creation, working on it,
thinking about it day in and day out in the above terms not to use that art-based language
when describing it to someone else. And yet, that language makes most business people
"nervous"; they do not understand it, and they do not want to hear about it when their
foremost daily concern is to create a product that must compete successfully in the
marketplace, that must sell well and earn profits. Our informant finds in that difference of
worldviews the reason why most good organizations can recognize the output of the
creative process and celebrate their creative people after the fact, while being not at all
comfortable with the processes. the systems. and the methodologies that the creative
people use.
One of the critical roles of the DCL then is to "mediate between the vocabularies and the
worldview and the processes of designers, and the vocabularies, the worldview, the
orientation and the processes of the business people." They play that role of mediator by
"lend[ing] a narrative, or creat[ing a narrative" around the work of the designers so that
the non-designers can understand what the designers have done. and why they have done
it in that particular way. Our informant explained that a critical element in this mediation
process involved explaining the historical progression of the designers' work. to help the
business people in the organization "understand the decision making process that's gone
on throughout the [design] process. so that they can make sense of what [the designers]
have created in their [own] terms." The next section describes a situation where the DCL
was called upon to play that role of providing a context for understanding the
development of a product.
An example of the DCL's internal translation and mediation roles
As an example of the DCL's role in "weaving a narrative" around an NDI design. our
informant described the development of a product that had been announced shortly befbre
our interview. The product in question is Nissan's entry into the small SV
189
(sport utility
vehicle) market segment, then mostly populated by Toyota's RAV4 and Honda's CR-V.
(Nissan's entry has since been introduced to the market as the Exterra, a product that
otffers some unusual features, such as a first aid kit built into the rear lift gate, and an
interior bicycle rack.) Since these competitors had already been on the market for some
time, Nissan was playing catch up; its project had to be done under an "exceedingly
compressed development time" in order for its entry to be viable. Nissan's design called
not only for a "low-priced" SUV but also for an "authentic" one. As our informant
explained, most SUVs sold today are not really used for the purposes for which they were
designed. namely. off-roading and sports activities; rather, they're used to go shopping
and as substitutes for regular cars and minivans. That fact was relied upon by Nissan's
competitors in designing their small SUVs: both the RAV4 and CR-V are based on
economy car mechanicals (the Honda Civic in the case of the latter). and do not have the
ruggedness and handling characteristics needed for off-roading. By contrast. Nissan's
"authentic" SUV is intended for people who would actually take it off-road, or use it in
their outdoors activities to carry their sports equipment.
Due to the radically shortened development time allotted to this product. it was not
possible for NDI. and especially NNA. to "solicit the kind of consumer input that the
product planners and marketers normally like to have." Specifically, there was no time to
expose the proposed designs to focus groups and consumer clinics along the way. in
order to reduce the uncertainty the planners feel about being on the right track. This
"system of checks" on which the corporation relies is based on straightforward thinking:
if a new product is aimed at a particular segment of the market. then. in order to assess
whether the company's proposed offering in that segment is likely to be well received.
company researchers go out and ask customers who already own various competitors'
products in that segment their opinion about the new design.
There was no time to conduct such research. and yet. "there was a great deal of
nervousness within the corporation as a whole" about the direction of the new design. To
address that concern, the DCL visited a few local areas where people engage in the kind
of activities targeted by the designers. Specifically. they went to mountain-biking trail
heads. they went to the local beaches to talk to surtfers. and they visited a nearby ski area
190
to talk to snow-boarders. During these visits, "they concentrated on the people and their
activity" without regard to the kind of vehicle they were driving at the time, be it a
station-wagon, a sedan, a sports car, or a sport-utility vehicle. They focused on the kind
of equipment these people brought with them, and "the issues involved in managing and
coping with that equipment." These visits were conducted at the outset, and they "helped
inform the designers." Seeing their target customers getting into their cars covered with
mud and sand and snow led to new ideas and developments such as unique fabric
coverings for the seats. The DCL researchers documented their work photographically
and created a collage and a narrative that they used in explaining the design decisions to
the planning and marketing executives. Our informant describes
.o then when all the marketers and the product planners came down [from
Los Angeles, where NNA is based]. and there was the vehicle that we
created,we couldsort of go back to a primary source and say. This is
what iformed our design eflbrt. We went and we spoke to these people.
We talked to them. We observed them. We documented them trying to cope
with these kinds of objecls. And consequently' we have incorporated X. Y
and Z/eatures into the vehicle. And so that was enough--that sort of
provided the link between the object and the observers. [...] Because the,
could see, Okay, these are the people: I can see who the people are: I can
see what their problems are. and I can see that you've addressed them
with these.features. So that was enough to get them to go ahead and bless
the design.
Our informant described this approach as simply relying on "common sense". He
complained about the difficulty of dealing with decision makers who seem to "need proof
beyond their own existential, experiential proof." Although the work of the DCL,
described above. did not provide the numbers and charts that some decision makers
normally associate with that type of proof. it did give present them with a believable
narrative that they could understand, accept, and rely on to move forward with the
project.
Later on in the development process. when the proposed vehicle was finally exposed to
potential customers who engage in the activities around which it was designed, they
received it "very favorably." Of course, it remains to be seen how economically
successful it will be over its lifetime.
191
Other examples of internal integration
The above example not only demonstrated the internal integration function of the DCL
but also its role in external integration, where it bridges the distance between the
designers at NDI and their target customers. This role will be discussed in greater detail
later in this section. Our informant provided other examples of DCL's internal integration
role, in which consumer information was not directly involved. One such example is the
case of a prototype for an electric vehicle that NDI worked on following the introduction
of California's Zero Emissions Vehicle regulations a few years ago (requirements that
have since been rescinded.) Our informant described a situation as one where there were
"fundamental differences in orientation" between the designers at NDI and the product
planners at NNA towards the electric vehicle project. The designers. "being optimistic,"
saw an opportunity to make a vehicle for the person who is really committed to reducing
pollution, while at the same time designing a "really cool and appealing" product. The
excitement of the designers was at odds with the planners' attitude. who felt harassed for
having to deal with "this damned regulation" Early on. the DCL, hosted a conference at
NDI to try to build consensus around the project. and they invited representatives of
different parts of the corporation. Our informant described that meeting as "very
contentious" and "a real dustup." No consensus was reached on that occasion: the product
planner in charge kept pushing for a sporty vehicle with a sloping hood. while the
designers kept disagreeing with him. The designers went on to pursue their own ideas.
Later on. when the planner saw their initial work. a prototype that no one at NDI thought
looked remotely like a sports car. he exclaimed "That's what I meant! That's what I
meant!" Although the DCL was not particularly successful in carrying out its role in this
case, our informant gave this example to underline the nature of its internal integration
role. The product planner and the designers basically had the same thing in mind all
along, but they talked about it in radically different languages. The product planner was
unable to verbally communicate what he could only vaguely envision as an "exciting"
and "cool" vehicle without falling back on his familiar vocabulary of"sporty" and
"sports car".
The other example offered by our informant also involves the verbal description of a
product. In this case, a product planner in the Nissan marketing organization in Los
192
Angeles sought the help of the DCL. At Nissan, not unlike other organizations, one of the
first steps in arriving at a product concept is the "OB stakes document", where "OB
stakes" stands for "out-of-bounds stakes". That document is intended to stake out an area
within which the designers can work, and it will generally answer such questions as what
kind of car they're aiming at, who the typical customer is expected to be, what role will it
play in the company's product line-up, how it might affect the overall product strategy,
andi so on. Typically, it will also describe what mechanical components such as chassis,
engine and transmission, the engineers expect to use in that product. The problem with
this process, as with any other design situation, is that in these early stages the planners
do not have a well-defined concept in mind: rather, the concept is still evolving. The
designers, however. need to start their work; in this case, they started exploring options
based on their understanding of the market in general and the particular segment targeted.
and on their own interpretation of the requirements. Not surprisingly, conflicts started
emerging between them and the product planners as to whether the options were within
the (still-evolving) OB stakes or not. It is at that point that our informant was called upon
to help the product planner develop some language that would better articulate the
concept in the planner's OB stakes document. Our informant characterized the role of the
DCL in this case as crucial in developing a document that would meet the needs of the
product planner. while allowing as much room for the designers to explore their various
options. He described the language he came up with as a "sort of cute haiku" using the
key words "emotion plus intelligence".
Ultimately however, the role of the DCL as described above is one of interpretation. It
cannot translate or interpret what is not there, nor can it through that process close the
gaps that were left open by the groups between which mediates. In this case, our
informant relayed that later on in the project, the fact that the product planners relied on
the DCL proved to be somewhat of a drawback, due to the lack of specificity in the
definition that the DCL had helped develop. Since the product planners themselves had
not narrowed their own understanding of the concept, they were not fully prepared to
give the designers more concrete direction when the latter reached a point in their work
where they needed it.
193
Again, our informant ascribed the dissonance that crops up between designers and
product planners to the different orientations of the two groups, the first being visually
oriented people by training and disposition, and the latter being more verbally and
certainly more quantitatively oriented. As a result of this orientation, the planners have a
harder time "reading and understanding objects, especially when they're in quarterscale"; "they cannot visualize something that does not exist yet." Our informant used an
analogy to the well-known argument concerning the definition of pornography to
describe what he believes to be the dilemma of the product planner who may have in
mind an idea that is not well-articulated: "I don't know what it is but I know it when I see
it."
The DCL and externalintegration
The other key function of the DCL is to provide a link between the organization and its
customers. Although it is often difficult to separate the two functions of internal and
external integration in product development, as we saw in the first example above, this
section will focus primarily on the DCL's particular approach to the problem of external
integration, of bridging the gap between the design organization and the people who
ultimately will decide whether to purchase its products or not.
In general, our informant was critical of the industry's generally accepted approaches to
customer research, starting with the reductionist practice of segmenting the market into
smaller and smaller niches based on product specifications and features.
The car industry breaks up the market: there's, what, now 400 and...
almost 500 different models of cars for sale in the United States. When car
industry people look at the market they break it up into segments. There's
a segmentfor sporty cars, and midsizesedans,and large cars.and station
wagons, and sport utility vehicles, and sports cars, and the whole bit. The
averageconsumerdoesn'tthink that way. They basicallysay, "I'vegot a
budget of $15,000 dollars. What can I get? I can get a truck. I can get a
coupe. I can get... "you know, whatever. And that's their starting point.
Although it may seem like a far stretch for a car buyer to be considering both a pickup
truck and a sporty coupe at the same time, that was exactly what the DCL personnel
194
found when they conducted some preliminary research for a new sports car. That research
is described in greater detail later.
According to our informant, the analytical approach to market classification simply does
not reflect the complex and non-obvious way in which consumers understand categories
and how they use them. More fundamentally, be questions the way in which researchers
typically measure consumer preferences for the various product characteristics and
features, which are the dimensions used in mapping out the space to be segmented.
Again, the objection is based on the significant discrepancy between the consumer's reallife experience and perceptions, and the analytical abstraction intended to represent that
experience; more specifically, between the complexity of a real-life car purchase and a
consumer focus group:
And what this [...] system would do would basically say, all right, let's
ignoreall that. And let's try and make thispseudo-scientific.And let's try
and get a pure take on-you know... The consumerscan sit around the
table and say, "Well,thefirst thing that mattersto me is value and
comfort. " But they're looking at a photograph. They can't get in the car.
They can't sit in it. They don't have Consumer Reports infront of them to
tell what the projected reliability is.
The focus of the DCL then, as the next examples will show, is to better capture, if not
understand, the complex real-world interaction of the consumer with the product, be it in
the context of the purchase decision or in the context of the day-to-day use of the vehicle;
and to communicate, if not explain, that complexity to the designers at NDI and to the
rest of the organization. Paradoxically, in so doing, it would seem that the DCL's work
has the effect of detracting from the work of the NDI's designers. Our informant talked
about the detrimental effect of focusing too much on a vehicle's external appearance, as
is commonly done during traditional consumer clinics and focus groups:
[...] By putting so much emphasis to the exclusion of every other
importantcomponentin their decisionmakingprocess,putting all that
emphasison the exterior appearanceof the vehicle,you do design a
disservice because then the design of the vehicle is carrying all the weight.
That the entire rest of the corporation in terms of their communication
strategy,their sales strategy,theirpricing strategy.the quality of their
manufacture,the qualityof their engineering,all that stuff, the burden is
195
being carried on what the thing looks like. And it gives the corporation a
distorted view qofwhat's important to consumers. And it also subverts
progressive design.
Such a statement may sound strange coming from a group whose purpose is to support
the designers' work; however, it is in line with the kind of design in which NDI sees itself
involved, namely, industrial design in an automotive context, as opposed to the more
traditional "Detroit style" automobile styling. It is worth noting that by helping reduce the
undue burden typically carried by the styling of the vehicle, the DCL is not simplifying
the work of NDI's designers; to the contrary. By providing the designers with a deeper
context about the customer's experience with her automobile, they give them more issues
to deal with in the design process, and they are add to the complexity of their task. The
following examples of external integration shed additional light on these issues.
DCL's approach
o consumer clinics: the case of the Maxima
The DCL's new approach to obtaining customer input into new product development
projects started out as a backlash against the traditional methods that were in general used
throughout the corporation, both in Japan and in Los Angeles (at NNA.) These methods
were analytical in nature. relying heavily on traditional design clinics and statistical
surveys, the results of which were accepted with little criticism and used mechanistically
in the development process. Our informant described that approach to the process of
soliciting customer input as "incredibly destructive to the output of our corporation in
terms of the kinds of cars that we make." He described the old process and contrasted it
to the new approach that NDI used for the Nissan Maxima, a model that had been
released less than two years prior to our visit.
In the old process, after the most promising design proposals had been developed into
quarter-scale models, a focus group consisting of owners of a similar type of car would
be assembled. During that focus group, normally run and moderated by an outside firm
that specializes in putting together and conducting such meetings, the participants would
be shown photographs of the various scale models. and they would be asked which they
liked best. Typically, the model that was liked best by the largest number of participants
would then be approved by the corporation for further development. The other candidates
196
would be discarded. And although most managers and executives in the corporation
thought that this approach was a "perfectly reasonable" way to obtain customer reaction
to a proposed model, our informant criticized it as "being utterly divorced from: a) the
way people think about cars, b) the way people experience cars, and c) the way people
shop for cars; plus, it does not make sense." One of the shortcomings of the traditional
approach, as he described it, is a direct result of the time delay between when the models
are shown to the focus group and when the actual product is scheduled to hit the market,
a gap of up to several years. He explained that it is "unreasonable to ask consumers what
their taste will be in four years" given that people's taste is so rooted in the present.
Typically, the same people who initially complain about all cars looking the same today,
when shown the photographs of designs intended to be "appropriately distinctive four
years from now" would find them inappropriate. During an earlier interview with the
designers of the new Altima, D1 talked about some of the problems with the conventional
approach from the perspective of the designer:
...when we used to just work with their [NMC in Los Angeles] research,
you know,they would take our quarterscale modelsand clinic them,and
they would come back with, you know: "Headlight was rated 20%;
unacceptable. Taillight, and this line, [the] door cut here, nobody [liked
it]... it rated badly... "And the whole car would be broken up into these
little analytic bits, and it just drove us crazy because we didn't know how
to relate to that. Ifyou take a customer, somebody who's not associated
with the design community, and put him on the spot and say, Do you like
that headlight?And he's lookingat the whole thing, and all of a sudden,
he's got tofocus on the headlightand he's got to make an evaluationof it.
Wellmostpeoplejust don't look at cars like that; they look at the totality
of it.
He also talked about the personal cost to the designer, of being subjected to such a
critique while observing one of these focus groups:
[The designerwould]sit behind a one way mirrorand watchthis, and
you're just like, Oh, boy! And you can see the executive group, they're
eating this up, andyou're watchingyour car, this designthat you worked
half a year on, go down the tube.
197
In the case of the Maxima, the DCL decided to conduct their own consumer research, and
to design it in a way to improve on the shortcomings of the traditional approach. Instead
of holding the focus group in the typical small room with the usual one-way mirror and a
hired moderator, they rented a converted power station for the occasion, a large hall that
was closer to a studio environment. Next, in an important break with the traditional
approach, the participants were told that they would be evaluating designs for Nissan.
Normally, the identity of the client firm is kept hidden from the participants (through the
use of the facilitating entity, for example) in order not to bias the responses. To the DCL
personnel, that consideration was outweighed by the fact that a consumer's reaction to a
new design cannot be divorced from the history of the product and the perception of the
brand. What may be seen as appropriate for a BMW may be viewed differently if the
brand concerned is Mitsubishi. for example. With that in mind, the DCL set up a "context
zone" where the evolution of the Maxima, through its three successive generations. was
displayed for the benefit of the focus group participants. Another significant departure
from the old practice is that the participants invited were owners of the different
generations of Maximas, not simply recent buyers of similar models. In addition, they
were screened carefully to ensure that they had owned their Maximas long enough, in
order to avoid the tendency among recent automobile buyers, who may still be
rationalizing their purchase, to react negatively to a newer design.
The focus group participants were shown the context zone, and the Nissan personnel
shard with them the history of the model. This was followed by a lengthy conversation
about the identity of the Maxima, what it communicates to them, how it differs from
competing models such as the Toyota Camry, and so on. Next came a phase during which
the participants were educated on the difference between looking at a real vehicle and
looking at a quarter-scale model of it. They were shown a quarter-scale version of the
then current Maxima, as well as the real car itself, and they were invited to go back and
forth between the two and to discuss the similarities and differences between the two
experiences. This prepared the participants for the final phase of the meeting, in which
they were shown the actual scale models of the different proposals on which the
designers had been working, as opposed to the standard photographic shots of these
198
models. This was an important factor in reaching the DCL's goal of getting more
constructive information for their designers:
... there was a dynamic that was set up in the previous methodology
where,by the way that the interactionwas structuredand by the types of
questions that were asked, almost by default you were inviting people to
critiquethe car. First of all they were shown three views:front threequarters,side, and rear three-quartersin photographs.And since the car
was represented in two dimensions those elements qf the car that came
forward were the graphic elements. So people would critique the wheels
and they would critique the door handles. They couldn't get a real sense of
the flavor of the car.
More generally, the DCL researchers made it clear to the participants that they were not
asking them to play the role of a jury passing judgment on, or grading and critiquing the
work of the designers. They explained to them that they wanted to "invite them into the
design process" and "to solicit their input" into that process:
We'dlike you to collaboratewith us in helpingus to understand what it is
that we've done. So that was a way of-we didn't want them to feel
compelled to criticize. We wanted to invite their input.
A critical first step in inviting the focus group participants into the design process. and
making them able to contribute meaningfully, is taking the time to educate them about
that process. To that end, the participants were told where the designers were in the
process; for example, they were told that there would be another set of quarter-scale
models, followed by a full-size model. Most importantly, they were told who the target
customer was, and the NDI personnel shared with them their thinking on the new model
and the company's intentions, information that is normally guarded very carefully by any
product development organization.
The following is an illustration of the kind of interaction the DCL researchers were
seeking to start with the participants:
We had four versions, bur differentmodels,four differentarticulations of
Maxima.And we shared with them what each designer'sintent was. So by
doing so what we were trying to get at was not whether you liked that
particularcar but do you understandit as a Maxima?Does it
communicate Maxima to you?
199
In trying to engage the participants in the design process at this deeper level, the DCL's
personnel were aware that they were placing greater demands and responsibility on the
participants than in the traditional focus groups. In fact, they had been warned by other
groups within the corporation that they would not be able to conduct such a discussion
with consumers because "they're not trained professionals." Our informant's answer to
this criticism was that if one were of the opinion that it is not possible to get "good"
information out of consumers because they are not trained in design, then why would one
bother to ask them in the first place. In fact, our informant described the responses that
the DCL got from their modified focus group as "incredibly insightful [and] articulate."
He gave the following example loose quotation from one of the participants who was able
to identify the particular mix of sportiness and luxury that is the hallmark of the Maxima:
Well, I understand that one and I really like the way it looks, but that one
is not a Maxima because it looks too luxurious. And the other one, Well,
that one looks too sporty, and I don't like it anyway. But that one in the
middle, I don't know, there's something about it; it's got the right balance
of luxury and performance attributes that is Maxima to me.
Our informant also described a typical reaction from one of the designers to the feedback
from the focus group:
'Wow, people understood what we did. This is how they saw what we did.
And, Yeah,I guess I neverthoughtabout that modelthat uay. hut now that
I look at it, I can see why that would signal luxury to someone whereas
this one signals performance.'
The above quotes highlight the open-ended nature of the exchange in which the
participants and the designers (even those not present at the focus group) were engaged.
During this conversation, the designers took the opportunity to explain themselves, their
thinking, and the resulting designs (sometimes indirectly, through the context zone and
the description of the target customer.) As a result, the feedback that they got from the
participants was more constructive compared to the criticism to which they had become
accustomed. They were getting intelligent descriptions of how the consumer understood
what they had done and why, and they even were getting some praise for their effort.
That kind of feedback helped to boost the designers' confidence in their work and the
200
overall project. That boost in confidence was described as being important to the work of
the designers at NDI, and it seemed at odds with the often-held view of designers as
being aloof, of working in a world of their own, and of being unconcerned about the
opinion of non-designers:
[...] our designersare not megalomaniacsand they often are... They have
their convictions,but they often are unsure:Did I do the right thing?Did I
go toofar? Did I notgo far enough?[...] in the old modelwhat came back
was critique and criticism, never praise: 'This is what's wrong with it.
This is what's wrong with it. 'And that wears on you ifyou hear it year in
and year out. So what it served to do within the building and within the
corporation at large is boost everybody's confidence around the project.
The open-ended quality that characterized NDI's non-traditional focus group was not
limited to the way it was conducted and to the conversation between the NDI personnel
and the focus-group participants; it also carried over to the way in which the information
that was obtained from the session was later used internally at NDI. Unlike in the past,
the results were not treated as a go/no-go decisions on specific design features or
elements. Our informant:
When we got the information out of that research effort, there was no
winneror loser.
What the information provided was a better understanding of the direction in which each
of the prototypes was likely to move the Maxima if that prototype were developed
further. That deeper understanding started a new set of internal conversations between the
different groups within the corporation, about the direction in which to take the next
generation Maxima, whether to evolve and refine the then current Maxima, or whether to
take the new car in a different direction.
Ironically, at the same time that more work and greater involvement was being asked
from the participating consumers, the responsibility for making the tough design
decisions was being brought back inside the organization. Our informant described that
change as follows:
So, in a way, what had happened in the past was that the corporation as a
whole was using the consumers and the consumer clinic really as proxies
201
for their decisionmakingresponsibilities.'Wow!No matter what I think,
we can't go with that model because the consumers didn't like it. 'So what
we did in this instance was, we got output that really returned the
responsibility to its rightful place, which is the management structure
that's supposed to be making the decisions. There was no winner. This one
means Maxima this. This one does this to Maxima. And this one does that
to Maxima. So then the discussions around the decision making process
were not about which one won the clinic, but about what is the proper
directionto take Maxima?And as a resultI think we made a better
decision than we would have otherwise.
This change was not merely a zero-sum shift in the locus of decision making, from the
voice of the customer (through the proxy of the focus group participants) back to the
design organization. The two parties were engaged in a joint process of developing a
deeper understanding of the issues and tradeoffs involved in these design decisions.
A second example of open-ended consumer interaction: the Altima
Our informant offered other examples of consumer research conducted by NDI which
was intended to be open-ended, in the sense that the consumer was not asked about his or
her preference regarding specific design features, but where the researchers engaged the
consumers in a free-flowing conversation around a very general topic. One case in point
involved research undertaken early on during the Altima project. The general aim of the
research was to inform the designers about different ways to think about car interiors.
The research was prompted by the observation that many people were using the cup
holders in their vehicles as "cell-phone holders", and by the observation that people were
spending more and more time inside their cars (something that is especially noticeable in
Southern California, with its hour-long bumper-to-bumper commutes.) Here is how our
informant described the stance of NDI's researchers as they prepared for this research:
... we brought in some Altima ownersjust to speak with them about their
car and talk with them informally. We didn't have a list of questions. And
what's interestingaboutdoingthat-see, traditionallythey wouldhave a
list of questions. 'I want to know this and this; here are 10 thingsthat I
want to know, that I have to get out of this group. ' Well. we want to know
what's important to the consumers.,not what's important to us. What's
important to them.
202
This approach led the DCL researchers and their respondents to engage in a "very rich
discussion about the notion of the car interior as a personal refuge." During this
conversation, while generally discussing the ways people use their cars, one participant,
an emergency room nurse, described one particularly bad day when, upon returning home
and finding her baby crying and her dog barking, decided to return to the garage and to
take a nap in her car. This led other participants to describe how they too often took naps
in their cars. Although this free-flowing conversation is interesting in itself as a method
for eliciting information, equally interesting and challenging is the way in which the
designers ultimately use this information. The designers are not expected necessarily to
act on it directly, by mechanistically translating it into specific product features. Instead,
this information is intended'to contribute to their general background knoviedge about
their customers and how they interact with their products. Our informant made that point
clear; referring again to the use of the car interior as a personal refuge:
Maybethat doesn't necessarilymean the designeris going to put a pillow
on the seat back, but psychologically, it informs them. 'Well, really, I
thought this was a machine for driving, but the consumer is thinking about
it as a refuge. 'So that leads you down a different avenue of exploration
thanyou might otherwisego.
A summary of the DCL's philosophy and approaches
A hallmark of the DCL is their flexibility. both in the range of tasks with which they are
called upon to help, and in the variety of approaches they use to carry out their work. The
DCL do not follow a given, fixed methodology, nor is their domain of action clearly
defined and circumscribed. For example, helping out the product planner in LA in
wording and refining his product brief-one of the examples presented earlier-is not
one of the DCL's 'official' tasks. Furthermore, in many if not all of the DCL's activities,
their internal and external integration roles are intertwined, as in the case, for example,
where they used an anthropological study of potential customer groups to explain the
designers' decisions to the marketing personnel. The DCL personnel do have a set of
principles or understandings that guide the way in which they approach any situation, and
these are the principles of interpretive thinking. One of these is an awareness of the role
that people's background knowledge plays in structuring their understanding and coloring
203
their thinking. That is the reason why, for example, the DCL held the conference
described above to kick-off the Maxima project, the conference that brought together the
product planners, marketers, designers, and other concerned parties, to answer the
question, "Well, do we know what a Maxima is?" In that conference, every participant
was asked to go off and write his or her answer to that question, based on his or her
beliefs and understandings. By doing so, the DCL personnel were making sure to start the
process of developing a shared understanding that would serve as "the building block" or
the starting point for the designers.
Another aspect of the DCL's interpretive thinking is the notion that additional
information or knowledge need not necessarily foreclose options or limit the designers'
range of exploration, as is the case in analytical definitions of information. Rather, the
additional knowledge has the potential of opening up new avenues and new spaces for
interpretation. In the case of the Maxima for example, the purpose of the meeting
described in the previous paragraph was not to capture the shared meaning of the Maxima
in order to hand it over to the designers and ask them to develop a new embodiment of
that concept. Our informant made it clear that although that was an option, the designers
were encouraged to interpret freely and to explore different directions:
Well, do we know what a Maxima is? And everyone went off and wrote it
down and, sure enough, yes, this was our shared understanding of what
Maxima is. And that was really the building block. And so the designers'
task was to take this shared understanding of whatever it means to be a
Maxima and create anotherone; or create differentvariationsof that.
Yet another element of the interpretive thinking at the DCL is the reliance on open-ended
questions and the avoidance of ready answers, and of questions that would lead to such
answers. That was covered in some detail in several of the previous examples, in
particular the general conversations undertaken early on for the design of the Altima
interior, and the more recent ones carried out in the context of the new sports car project.
As a related matter, the use by the DCL of research methods usually associated with
ethnography is another example of its interpretive bent (Geertz, 1973; Van Maanen,
204
1988).57 Such methods were described earlier, in the context of the research conducted
among surfers and mountain bikers.
An important element of interpretive thinking at the DCL, also discussed earlier, is its
rejection of rigid, well-defined categories that are based on analytical mappings from sets
of features, and its preference for more natural classifications that better reflect the reallife experience of consumers. This approach to categories follows the thinking of such
researchers as Rosch, and Lakoff and Johnson (Rosch & Lloyd, 1978; Lakoff et al., 1980;
Lakoff, 1987).
A clear expression of the DCL's interpretive character is its reliance on different
approaches to their work, as the context and the circumstances suggest or require. That is
clear from the range of examples presented above. The members of the DCL do not
believe in following one single approach or a fixed research technique all the time. This
however turns out to be a source of difficulty for the DCL in its work with the rest of the
organization, where people tend to prefer a given system or methodology that they can
apply universally and repeatedly, and tend to be uncomfortable with the ambiguity that is
inherent in such context-dependent approaches. Our informant's description:
So what we've been doing in activities like this as well is trying to sort of
inoculate the host corporation, ifyou will, with more progressive
methodologies. And we're always quick to say, Well look, just because we
did it this way under these circumstances does not necessarily mean that
the next time we have a different kind ofproduct that we're going to do it
the same way. We're going to try and develop methodologies that are
appropriate to the task and appropriate to the product. Which drives
people crazy because they want a system and they want to be able to just
implement something at the drop of a hat. Whereas we try to be much
more sensitiveto the particularitiesof a givenproduct and a given
program.
This openness to the particularities of the given situation and its context, and the
readiness to develop new approaches as needed are hallmarks of the interpretive approach
of the DCL
57
Clearly, this does not refer to the now obsolete notion of ethnographic fieldwork as scientific
observation, but to the more modern view of fieldwork as an interpretive process (Van Maanen, 1988,
205
The DCL is an example of an openly interpretive entity within NDI. Its members are
trained in disciplines-linguistics and art history and criticism-that give them facility
with the language of interpretative thinking, something that was obvious during the
interview with our DCL informant. Furthermore, one aspect of their job is to evangelize,
to spread that type of thinking throughout the organization; as a result, they are explicit
when they talk about interpretive issues and about the differences between interpretive
and non-interpretive approaches. In contrast, the next section deals with a group within
NDI that relies on interpretive approaches in a more implicit way, almost by default, due
to the nature of their work.
The Color Studio
NDI's approach to design, which NDI's president referred to on many occasions during
our visit as an "intuitive" approach, is typified by the work carried out at the Color
Studio. The Color Studio consists of four designers who are not only responsible for the
exterior colors on specific Nissan models, but for developing interior colors and fabrics
as well. NDI's color studio was set up relatively recently; at the time of our visit, it was
only seven years old. Initially, the studio supported the other NDI designers in their work
such as developing colors for the all-important quarter-scale models used in design
decision making, as described earlier. The color studio also undertook "minor change
programs", that is, fine-tuning the color offerings of existing U.S. models. The Altima
project was the first one for which the color studio was charged with developing an entire
color palette for a new model. That work started four years before the introduction date of
the car (as mentioned earlier, this introduction coincided with our visit.) Before the
Altima program, the major color design work for all of Nissan's products was carried out
in Japan, with input from the marketing organizations of the various markets.
The four color studio designers are all young women with academic training in the fine
arts. One of them has a background in fabric design; another in product design. They
interact most often with the designers from the other NDI studios. They also interact with
pp. 34, 44).
206
their counterparts in Europe and Japan, as well as the assembly plants in the U.S., where
the colors they design are ultimately applied to the product.
The color studio is a particularly interesting study site because of the highly intuitive
nature of the work carried out there. We interviewed the chief designer in charge of that
studio, whom we will refer to here as B, at length. She characterized the work of her
studio as follows:
... we work in really abstract sort of levels because color is so difficult to
communicate; and we deal in, I think, in a much different level than the
[other]people here. I don 't know if it's higheror lower,but it is definitely
a different level.
The very nature of working with colors, fabrics and textures, and the difficulty faced by
color and fabric designers in discussing and explaining their ideas and concepts to
outsiders, have resulted in B and her team developing certain unusual approaches to
communication and integration. These approaches are intertwined with B's and her
team's particular philosophy and beliefs concerning the way in which designers select
colors, and how these in turn become accepted by the public. Their philosophy is
generally at odds with the approach that is more typical in the industry. She underlined
one aspect of that difference in rather strong terms:
I think color designersand color design studiosattempt to act like they
know what [colors will be popular in thefuture]... like they have a crystal
ball and they know exactly... And we, I think, decided to say, You know,
we 're not going to pretend that we do.
Later, another aspect of that philosophy came through as she was commenting on the use
of surveys and statistics in data-driven design:
I can 't get a piece of 8 by 11paper with a chart on it and expectthat to
inspire me in any way and/or make me think beyond... Pfffit! I can 't do it.
Not only does she reject the idea that new colors simply jump out of the mind of the
individual color designer, as well as the notion that they can be arrived at by looking at
statistics of what people are buying. She also rejects the commonly held belief that
automobile colors follow, or are interpolations or extrapolations of, trends in fashion
apparel or other such "identifiers":
207
... we 're not going to say that there 's an identifier. And the biggest thing
that wefeel pretty stronglyabout is [that] fashion isn 't [it]. You don't
open a magazine and say, Ooh, there 's a purple, therefore it's going to
work on a car. Itjust isn 't... if it were that linear, anyone can do it; and
we don 't believe anyone can just, like, put a color on a car.
Rather, automobile colors are the result of a complex interaction between, among others,
the designer's intent, the forms and shapes that they developed or had to accept as a result
of some constraint, and the possibilities afforded by the available paint technology:
It has much more to do with the communication of what the designer on
that exterior[intended],and theforms, and thepaint technologies... t 'sa
lot more than just a hue to us. Too many people look at it as, it 's kind of
fashion oriented; it s not quite that direct.
The picture that emerges from this series of statements is that the role of the color
designers involves complex and intensive interactions and communication with the form
designers, the consumers with whom the former are trying to communicate, the paint
technologists, and so on.
This section will focus on the communication and integration approaches and
mechanisms that the color studio designers use in the course of conducting their work,
which requires them to exchange highly abstract and intuitive information. As in previous
sections, a distinction is made between mechanisms of internal integration and those of
external integration. As explained before, however, that distinction is sometimes an
artificial one, as internal discussions often relate to the different team members'
perceptions or understandings of the customers' wants, needs, and desires, without the
customers being directly involved in these discussions. Still, the distinction is a useful
one for the purposes of organizing the presentation.
Examplesof internalintegration
The Nissan global color meeting
The 1998 Nissan global color meeting provides a number of examples of the color
designers approach to internal integration. Once a year, a meeting is held in Japan to
decide on the colors offered on Nissan cars for the coming year. That meeting brings
together Nissan color designers from the US, Europe, and Japan. The reason why colors
208
offered around the world need to be coordinated is the limited capacity in Nissan's plants
for the number of colors that can be applied to cars, combined with the differing regional
tastes and preferences.5 8 Our informant explained:
They [the plants] just can 't deal with twenty colors for America, twenty
colorsfor Europe and twentycolorsfor Japan, so there'sa lot of arm
wrestling, you 'd call it, andjust--This is what our market needs...
Going into that meeting, the NDI color designers took with them a set of"questions" that
they believed would set the theme for that year's meeting, and raise a set of issues with
which they had been contending. Each question was embodied in a display that was
intended to pose the question and suggest ways of thinking about it and discussing it.
According to our informant, she and the other color designers at NDI "knew that we had
to make a statement in Japan," so they decided to focus on the "biggest questions" that
relate to their work and to that of the other color designers at Nissan.
The first question was referred to as "Redefinition", and it dealt with the public's
perception of colors, and the color designers' attempt to understand and redefine how that
perception works. Unfortunately, our informant could not tell us more about that question
or the display, for reasons of confidentiality.
The second display was entitled "Acceptance". The question that the designers were
attempting to address here was the public's perception of softness in seat cloth, a
situation that was described as a "continual problem" by our informant. Consumers in the
U.S. tend to prefer the softness of velour-like fabrics, a type of construction that is very
uncharacteristic of Europe, where people favor a flat, very structured, "modernist" type
of construction. The problem is that "most designers despise this very American, very
fluffy, pile-like stuff." The display was intended to help the designers better understand
the acceptance factors that come into play in the mind of the customer. To that end, the
display they assembled brought together samples that convey not only different levels of
softness and hardness, such as samples representing different types of fabric as described
58
Our informant indicated that Japanese and US consumers tend to be much closer in their color
preferences, compared to the European ones. The latter favor colors with high levels of "chroma", that is,
bright vivid colors. The former prefer colors that are more subdued. Our informant could not decide
whether the US and Japanese were "a little more refined" when it came to colors, or whether the
Europeans were ahead.
209
above, but also disparate items that inform such notions as "hard" and "soft", or that
inspire in people such ideas. For example, the display contained a marshmallow on a
stick and a piece of stone. Like the previous question, this exercise was also one of
redefinition in a sense, a redefinition of what is a soft and what is a hard fabric. The
discussions and conversations that were prompted by the question and the display that
went with it resulted in some new ideas for a fabric that "mixes, that blurs the boundaries
of what it is, but that allows us to move, maybe as a segue, to the direction we'd like to
see things going."
In this example, as in many others, it is notable that the new design or the new idea was
not the result of researching and mapping the customers' perceptions and preferences,
and then targeting the median or some other location in the taste or preference profile.
Clearly, the whole exercise was prompted by the state of customers' preferences;
however, the statement quoted above points more to a "quasi-activist" designer, who is
attempting to steer the consuming public in a direction that she likes, that she believes
people would find interesting and appealing and therefore ought to turn to. But she is
doing it in a graduated fashion, in a creative way that smoothes the transition she foresees
the customer will have to go through.
Not all the questions and displays related to highly subjective customers' perceptions.
The third question, for example, dealt with the technical issue of paint quality.5 9 The
problem in this case was the result of differing environmental regulations in the US and
Japan, resulting in different paint quality when the identical paint stock is applied in
different plants. Again, this issue was not the result of customer complaints or complaints
from the plant, but rather by the designers themselves being unhappy with the results they
were seeing: "We as designers were frustrated by it." The problem was very noticeable to
them when different models, manufactured at different plants, but both painted with the
same color code, were viewed side by side. This display consisted of ideas for different
surface treatments, with objects that had polished surfaces and sandblasted surfaces, or
59
US plants are required to use high-solids paints, that is, paints containing a lower proportion of thinners
or solvents that will ultimately evaporate into the atmosphere. As a result, the micas or aluminium chips
in metallic paints lay at a different angle, resulting in different levels of reflectivity and giving the
painted surface a different look. One solution to this problem would be to use a water-based paint
process, but that would involve large capital investments.
210
that alternated between the two on different sides. Here again, the question resulted in
new ideas for finishes that would bypass the problem. Our informant could not reveal the
details of the new finishes, which are expected to appear on the market in the next few
years, but she explained the approach as follows:
We 're saying, Ok, fine, if we can 't do what we want, how can we do what
we don 't want well; meaning, maybe solid colors, things that outsmart the
system in a sense.
The final one of the seven displays looked like Batman's cape. It was called "Tension",
and it consisted of a piece of rubber that was mounted in the middle of the frame of the
display, stretched in many directions by strings attached to the periphery of the frame.
This display was intended as a "visual metaphor of how it felt" to the NDI color
designers working on so many different things. Our informant explained:
... as a color designer, you 're switching gears so often and it 's so
frustrating that we said,you know,wejust want to [...] to admitto it and
tell everyone that we are under an extreme amount of tension; we are
running too manyprograms,too many thingsall at once, and itfeels this
way.
Our informant also explained that this was "purely an emotional thing," that they, as a
studio, were dealing with this issue at an "emotional and personal level." One could
imagine a more analytical way in which they could have approached that problem; for
example, they could have listed the projects and tasks for which they were responsible,
estimated the resulting workload on their studio, and made their case for more resources
or a reduced workload. Instead, they were thinking in terms of indirectly asking for help,
by talking about their workload openly and accepting that designers need not always "act
like the Rock of Gibraltar."
... we thought,by just ignoringthese really big questionswe're not getting
anywhere with it. Let's acknowledge that there are questions and we can't
answer them, unless we try to take this huge thing, as a group and deal
with it and admit that we don't have--youcan do words allyou want,you
know, Oh yeah, well, this is... But we said, let's come up with--we're not
sure how we're going to answer this. We don't know how to screw the
perception of the public. We don't know how to get quality. We weren't
sure. These were just questions that kept floating around us, and then
211
when we started working on the progress, we said, Wait a minute, the
quality, when we can't get it, what can we do? And then the answers came.
And, it's the same up here. Okay, but what can we do? What, [inaudible]
different with the yarn over here...
Applying colors and textures to forms that don't exist
The previous section on Nissan's global color meeting provided examples of interpretive
practices of internal integration between geographically dispersed yet functionally similar
groups (namely, color and fabric designers from the US, Japan, and Europe.) In this
section, we look at the internal integration role that the Color Studio members play within
NDI itself, in the course of their interactions with the exterior and interior designers.
When NDI is tasked with developing a new model, the work of the Color Studio begins
immediately. Our informant, B, and her team have to start developing a color palette for
the exterior, and fabrics and colors for the interior, from the start of the project, before
any exterior shapes or interior forms have been sketched. This early start is necessary due
to the long lead-time required to develop colors and fabrics. In order to insure the
durability and fade resistance of every exterior paint color and every interior fabric,
extensive and lengthy wear and exposure testing must be carried out. In effect, as our
informant put it, "The color happens before the form; in most cases." At this early point
in the project, the Color Studio, just like the studios responsible for the exterior shape and
for the interior design, is working from the brief given to them by the marketing
organization (the "outer bounds" document described earlier in the DCL section.)
During the interview, it became clear that our informant found this situation problematic
for two related reasons. First, as discussed at the beginning of the Color Studio section,
she is a highly visual thinker who needs more than a "just a piece of paper" to "inspire"
her. She describes the typical product brief as follows:
It's really weak,I can tellyou. It's one of thesethingsthat [makesyou
say], So what? I know that they're 24 years old and mostly skewed male
and graduatedfrom college. That doesn't really tell me a lot.
Second, she believes in a holistic approach to form, color, and texture. In the case of
exterior color for example, this was clear from her criticism of a show car displayed by a
212
competitor at the Detroit automobile show the previous year. That car was painted with a
novel type of paint that changed color from green to red depending on the incidence of
the light on its surface:
It looks like the car was beingdone over here, and the color design was
over there, and they didn't have the sameplan infront of them of what it
was going to be. [...] I don't think they ever lookedat that color, any of
those colorson that car, and made sure and said, Yes, this is where we
want to launch it because it's appropriate on this car.
And in the case of the interior:
It's really hard for me tojust slap a materialonto a form, and bring it...
make it work. Which is basically, I think how too often cars are designed;
as opposedto saying, Okay,I want this-maybe it's the crash pad, the IP
[instrumentpanel]-but couldwe do somethingdifferentthanjust slap
vinyl on it? And you have to design thatfrom the beginning.I can't do that
at the end. It's very difficult. [...] The shape and the materials are the
same.
Designing the interior fabrics and colors is even more difficult and complex than the
exterior color palette because the exterior shape affects the whole interior environment.
In particular, the shape and size of the greenhouse (the glass area) and the disposition of
the roof pillars can change the feel of the interior, and therefore how particular colors and
textures will look in it.
In effect, the tasks of the three design groups-exterior, interior, and color-are
highly
dependent on one another. Furthermore, the interdependencies are difficult to articulate
or to describe in analytical terms.60 For example, when pushed to explain what she did
not like about the competitor's show car mentioned above, our informant answered: "It
was too much. It did not have to go from red to green." Clearly, she was basing her
assessment on an overall impression, not on a specific analysis. In fact, when one
somewhat technical explanation was suggested. that perhaps the ever changing color
60 Referring back to the section on the Design Structure Matrix in Chapter III, it is interesting to note that
the interdependencies described here are, to a great extent, idiosyncratic. A color designer other than our
informant, one who may be comfortable developing a color palette in the abstract, without working
directly with the specific shape the color is being applied to, or using a generic form instead, would not
encounter the same interdependency.
213
made it difficult to accurately perceive the angles and curvatures of the body shape, our
informant did not comment on or pursue that line of reasoning.
One way in which B and her team deal with the circularity of this situation is through
extensive communication and "a lot of discussions with the designers" in the other
studios at NDI. Most often, these discussions center around highly visual elements and
concepts, and B and her colleagues rely on displays boxes that contain assemblages of
objects, similar to the displays they used for the global color meeting in Japan. These
boxes are assembled ad-hoc, to address particular issues or to make particular points. Our
informant uses these assemblages in a number of dialogic exchanges: a personal dialogue
where she is trying to clarify and organize her own thinking on the issue, and to explore
different dimensions and new directions relating to it; a dialogue within the Color Studio,
where she discusses these issues with her colleagues; and, of course, in the dialogues with
the other NDI designers.
One such box, the "potato display", was developed by B to explore different "levels of
thinking [on] how we deal with materials as humans." The display box contained models
of a potato in its natural state, of mashed potatoes, and of French fries. It also contained
some of the implements used to process the potato from one state to the other, such as a
potato peeler and a cutter. B described the display and her personal thinking and
motivation for putting it together as follows:
We start with this very natural thing. and then we start manipulating it;
and then we manipulate it more; and then we vermanipulate it; and then
we... It circulates almost back to this, to me. And perceptions; I was
dealingwith perceptions.Theseare like real andfake; and we do this
real-fake thing that isn't necessarily... I think that's where we are in terms
of automobilesright now:we do thisfake animalgrain and we do thisfake
wood. We throw in real leather. There's this mixture of weird things to me.
And I thought, okay, after that, we go to this level, where we take...
This particular approach, using the different forms of potato as a model, "was not
panning out" for our informant. She admits that although that particular display could
help her make her point, She felt that she needed to make her point using colors and
materials "that would matter to the designers." Perhaps the potato display was too
214
abstract, even in the NDI environment. In any case, it played an interim role in the overall
conversation about material naturalness.
The display B developed next contained several different materials-plastics, rubber-in
their raw or natural state, whatever that "natural" state was. In some cases, it was the state
of the material as it occurred naturally, but in several cases of man-made materials, the
state was one of minimal processing. For example, a few items in the display looked like
blocks of latex similar to that used in medical gloves and similar devices. The display
contained some colored materials as well. Again, in the process of choosing the materials
for this display, our informant was going through an interpretive process with herself:
I was doing it because it was something that I had to get out of my head. It
was something that I had to make dimensional to remind myself... When
you have too many things floating around, it's so easy to think, Natural,
organic, pop, colorful, or lack of color, clear. Or, you know... When you
start, sometimes you start dealing that way, very machine like... I kept
feeling confinedby that, and saying, Okay,how can I get out of this spiral
I was in. [...] I decidedI didn't want to do anything unnatural.Ijust
eliminatedthatfake animalgrain and this and that and, maybe,there's
other opportunities. So then I said, OK, the next one might be actually
where we take this very raw materials; they're not designs. These are
thingsyou wouldget... you know,it 's the inherentcolor of the materialI
guess is what I'm saying; without designs.
Clearly, B had a certain goal she was trying to achieve through this interaction, and she
had certain ideas in mind, which she was trying to develop and to communicate. She
describes her objective as follows:
I was trying to influence the designers to think differently about, when and
how they choose a form, to think about how, eventually, what material it
might be
Referring back to the typology developed in Chapter III, this would tend to place this
interaction close to the pragmatic interpretation type. The instrumentality of pragmatic
interpretation is clear in that she wanted to influence the thinking of the designers. She
wanted to get them to move away from what she saw as a strange mixture of materials
and textures. She also had a sense that there was some type of cyclical evolution that
materials underwent, and that it may be possible to move on to the next phase in that
215
evolution. Yet, her ideas were ambiguous, not fully formed nor clearly articulated, as is
clear from her descriptions above. By presenting and explaining the displays to the
interior designers, she was not proposing that they use specific designs or textures. Going
from different forms of potato to a material for an automobile interior would obviously
involve more than a trivial mechanistic transformation; similarly for the various forms of
raw plastic and latex. She was presenting her abstract ideas, along with the contextual
material contained in the displays, to start an interpretive process, a set of conversations
between her and the interior designers, but also between the interior designers and their
own work, with the purpose that this directed interpretive process would then potentially
lead to novel materials for car interiors.
During the visit to the Color Studio, several similar examples of highly interpretive but
directed interactions between B and her colleagues on the one hand, and the designers in
the other studios on the other, were discussed. One of these revolved around the design of
steering column control stalks and their texture. Through chance, our informant had come
across some interesting woodworking tools that had what she thought was an interesting
knurled handle. As a result, she became interested in "things that were knurled or
milled", and she started thinking about ways in which one could come up with "more
interesting stalks" as opposed to reusing the same carry-over stalks in all their cars. She
decided to communicate her ideas to the designers, and to try to move them to think
differently about control stalks and, more specifically, to think in terms of knurled
designs. To that end, she put together an assemblage of different tools and objects with
differently knurled surfaces, one of which was an exquisitely machined miniature
hammer with a knurled metal handle (the very object she had discovered earlier.
A similar example involved another member of the Color Studio, herein referred to as J.
In that case, J was thinking about the evolution of masses in shapes and forms, and she
was trying to develop a vocabulary for discussing hard forms and soft forms with the
interior designers, and use it to suggest materials and patterns to them. She developed a
display using different materials in different phases-solid, liquid, and gaseous. The
display included pieces of rock with sharply defined edges and angles, to rocks with
softer forms. Also included in the display were small containers of liquid, some with
216
objects moving in them, as J had become fascinated with fish and things in motion in
water. (At that point, she also started looking into fishing and fishing gear.) The display
included images of liquids beginning to boil, with different stages of bubble formation, as
an intermediate phase between liquids and gases, again looking for ways to convey
different, novel ideas about textures.
All these examples are characterized by a known direction, a certain goal, but one where
the ultimate embodiment of that goal is unclear and yet to be articulated. Neither J nor
our respondent was trying to suggest a specific design or shape or texture to the
designers. In fact, the open-endedness of the concepts and ideas that were presented in
the displays was such that our informant had a rather difficult time repeating, or
explaining in her own words, how her colleague presented the contents of her display to
the interior designers. Clearly, the contents of the displays by itself, and whatever ideas
attach to these objects at any one point in time, were much less important than the role
they played as a starter used to initiate an exchange of ideas.
Examplesof externalintegration
The Color Studio at NDI provided some of the most compelling examples of highly
interpretive approaches to external integration. Our informant described several situations
that illustrate the Color Studio's unusual approach to understanding their potential
customers. Three of these are reported below. The first example, briefly mentioned in the
Introduction (Chapter I), deals with understanding the buyers of sporty cars. The other
situations were not directly connected to a specific car project.
"Market segmentation" by bag of "chips": Understanding sports car buyers
A short time prior to our visit, NDI had been asked to start working on a new sports car
design program. The purpose of the program was to develop a new Nissan Z car, a
successor to a line of highly regarded and successful sports coupes stretching back to the
original 240Z model of the early 1970s. The last Z model, the 300ZX introduced in the
late 1980s, met with much success. A few years later, however, it had fallen victim to the
unexpected collapse of that segment in the sports car market, along with its competitors
in that segment-the Mazda RX-7, the Toyota Supra, and even the Porsche 944/968. As a
217
result of poor sales, all these models were ultimately discontinued, at least in the US
market, even though replacement models had not been developed. A number of factors
had been advanced as contributing to that unexpected change in the market, including
demographic shifts and a change in people's behavior from the conspicuous consumption
of the late 1980s to the more sober early 1990s. One of the explanations held that people
who in the past had opted for sports cars had in recent years replaced them with sport
utility vehicles and pick-up trucks, for a variety of reasons.6' On the other hand some data
from the marketing organization also suggested that the converse was also true, that in
one survey, half of the customers who ended up buying sports cars were initially
considering trucks and SUVs as well. Jerry Hirshberg, NDI's president, explained:
Hey, guys, let's check our own intuitions here. Do we really know what a
sports car is? Is there any new language? Are there any new references?
Who are these people? I'm already intrigued that people went in to buy a
truck and they'regettingsports cars. What'sthat all about?Half these
people wanted to buy trucks and walked out with a sports car. Now no
marketinggroup is goingto discoverthat. That'snota questionyou think
to ask before. " What truck were you interested in? " We may be aiming for
the entire wrong market.
The Color Studio was involved in the research project aimed at developing a better
understanding of who these sports car buyers were, and what their interests were. Our
informant in the Color Studio used different language to convey the same reasoning:
We decided... We said... We 're working on this program, and [we] said,
Who are these people? We really couldn't decide who they were and why
they would want to buy this product. And yet, it 's always easy to put them
in a market segment that... we think that it 's changing, and that people
might pullfrom a different segment...
To help them shed some light on this question, our informant in the Color Studio and her
colleagues devised an unusual approach to consumer research. The normal approach for
this type of research would have been to hold a focus group. This involves inviting a
group of people from the relevant demographic group, sitting them around a table, and
61
It should be pointed out that knowing the real reasons for the change in question is not essential tbr the
description of the practices in this section.
218
having a professionally trained facilitator conduct a conversation with them about who
they are and what they do. B described the alternative approach as a way "to let people
have fun with research." She and her colleagues assembled a collection of forty small
images, each picturing some item that they deemed "interesting". The images or
photographs were on "chips" measuring approximately one by two inches. They covered
a wide range of topics, from food-a picture of a hamburger and fries, a chicken, an Aunt
Jemima pancake syrup bottle-to clothing or fashion accessories-watches, sunglasses-to pieces of furniture-a Louis XIV armoire to a modern chair. Our informant explained:
So I was trying to have some design classics, some sort of more modern
things, vintage...
The above description and statement might suggest that the selection process for the
images in question was methodical, perhaps even following some experiment design
principles. That would be misleading. One of the images our informant included is an
image of a baby duck; she described her reason for including it as follows:
You know what, when we did it, we didn't like over-strategize or think
about the images. Like, I insisted this be here, and everyone said, " Why? "
And I said, "I don't know,Ijust have a feeling. Don't ask me why." [...]
It's a chicken, it's a duck,yeah, it's a duck.It's cute. [...] It 'sa baby duck
It is notable that out of the forty images in the set, only two or three represented cars,
even though the purpose of the whole exercise was to understand car buyers.
Furthermore, the car images they picked included a model that is not sold in the US, the
Renault Twingo (a one-box mini or city car, noted for its unusual interior design.)
Having decided on the set of images, our informant and her colleagues put together
identical packages that consisted of the chips, a few sheets of paper on which to glue
them, each sheet corresponding to one of three categories--"Like", "Don't Care", and
"Dislike". They included a set of instructions, along with a return envelope. Those
recipients who chose to complete and return this unusual questionnaire to include contact
information, and to indicate whether they would be interested in participating in a
"roundtable" discussion (in case our informant and her colleagues found them
"interesting").
219
The process by which potential participants were selected is no less unusual than the
design of the research instrument itself. Our informant, still in apparent disbelief herself,
described the process thus:
We went around and had a little envelope... We drove around... What we
did is, we put these on... We drove around... Can you believe this? We're
nuts... Weput these on people's cars that were competition [sports] cars,
in a sense... It's cool, huh?
The questionnaires were dropped off in shopping mall parking lots in several locations
around the country, including such disparate locations as San Diego, where NDI is
located, and South Carolina. Every car that received a questionnaire was also
photographed, so that the Color Studio personnel would have a record of what that
participant's car looked like when they selected it.
Finally, every participant who mailed back a response was eligible to receive a gift
certificate for a store of his or her choice, the available choices including a record store, a
bookstore, a home furnishings store, and store that sells fashion accessories. The
participant's choice was used by the NDI researchers as another indicator or "data point"
about the participant, albeit in a highly scientific way (as we discussed below.)
At the time of our visit, our informant and her colleagues at the Color Studio were in the
process of organizing the responses they had received, and it was clear that their
approach to sorting through the responses and extracting any knowledge from them
would be as non-analytical, as "intuitive" and interpretive as the process through which
the grab bag of image chips was assembled. The images were divided among the four
members of the Color Studio, with each responsible for sorting and organizing the
responses corresponding to ten of the forty chips, and for interpreting the results. Our
informant, for example, was tracking the baby duck mentioned earlier. As it turned out,
all the participants expressed a liking for that image. In discussing the implications or
meaning of that outcome, it was clear that she had not decided, a priori, what a positive
response to the duck would indicate about the participant. Rather, she was in the process
of developing an interpretation for that response after the fact:
220
Everyone likes the baby duck, so I don't really know what that means; I'm
still workingon tai.
And:
There'sa vulnerability,I think,or a cutenessor babylike...or
awkwardness, or... I'm not sure. These are all questions that I have that
I'm going to just pose to other people. Why do you think the duck's here
[in that category]? And it opens, maybe, communication?
Clearly, she was not following some algorithm or set of rules to translate a participant's
responses and demographic information into specific design preferences, such as, "If the
participant likes the baby duck, and if the participant is a male, and if his or her current
car is not black, then that respondent prefers bright colors such as yellow or green." In
that sense she was not following a straightforward pragmatic scheme, in which a
particular response, combined with certain contextual elements, result in a particular
understanding.
The above description points to a process that is more open-ended, more ill-defined a
priori, one that is best described as one of hermeneutic interpretation. By including the
baby duck (and several of the other images as well), our informant was starting a set of
open-ended conversations, between her team and the people whose cars were selected,
but also between herself and her colleagues (as the last sentence of the above quote
makes clear.) Other than the loose direction of generally leading to a better understanding
of sports car buyers, these conversations did not appear intended to provide direct
answers to clearly formulated, well-planned questions. At the beginning of the
conversation, the question posed by the baby duck picture was not clear in our
informant's mind: she thought the image was interesting, but she did not have a clearly
articulated reason for including the duck. Consequently, the responses received could not
be expected to have a clear and unambiguous meaning within them. The quote above
indicates that whatever meaning these responses will ultimately carry will emerge from
the conversations and interpretive processes in which our informant and her colleagues
were entering into.
221
It should be noted that our informant's statement does not seem to point to a carefully
planned set of conversations. Her last comment, "It opens, maybe, communication".
suggests that she did not look at the conversations that would result from this research
exercise in an instrumental way. She does not state that her objective is to have a set of
conversations, and that this research is the means for initiating them. Such a stance would
be at odds with the highly interpretive mindset that imbues the thinking and most of the
activities of the Color Studio.
It is remarkable that our informant and her colleagues seemed at ease living with the
ambiguity and lack of precise definitions that are inherent in their approach. In that
respect, they are different from many design teams who feel the need to fully define what
they expect to learn from any research exercise as they are designing it, and to crisply
categorize and quantify the data they obtain from the field. When asked whether she was
concerned that her research design might have been biased, and probably did not come
close to meeting accepted standards for sampling or design of experiments (for the
selection of the chips. for instance), our informant answered: "We don't care." And later.
when it was pointed out to her that some of the responses were not understandable, her
answer was that the point of the exercise was not to make sense of all the responses. That
exchange is transcribed below:
Interviewer: This is something I don't understand...
Informant: I think. It doesn't matter. You don't have to. Maybe that's... I
think our biggest thing is, if we could.figure out why the thing was there
for every object, then you lost it...
Interviewer: Because...
Informant: I don 't know why.
This would indicate again that our informant understood that she and her colleagues were
not seeking answers to specific questions, but something that could not, or rather, should
not be easily articulated and captured through an analytical framework. What they were
looking for is an interpretation, and specifically their interpretation, of the consumer's
mindset, an interpretation that would then illuminate their work in a non-obvious or
222
straightforward way. The picture that emerges here is one where the Color Studio
designers, or our informant at least, accept as a premise that people, their thinking and
their preferences are very complex and difficult to capture in words and numbers; at least
that part of their thinking and preferences that have to do with colors, shapes, and
textures. And that the best way to understand them is to accept them as a black box, and
to train one's own "black box" through exercises such as the one described in this
section. An apt metaphor for this situation is that of a neural network that is being trained
to give an appropriate output (in the form of a design) for a set of inputs (a set of
observations about potential customers and the world at large) (McClelland et al.. 1988).
The exercise can be thought of as a training data set, with the "'neural net" parameters in
the mind of the designer being modified and adjusted through the observations and
conversations described above. This metaphor can be glimpsed in the following statement
by our informant:
You know, it opens tip your mind of maybe how people might be thinking.
And what swasinteresting too, how I started even looking at the
information, is taking, like these sunglasses, and saying, okay. looking at
the name and the age, and say, okay, where do I think that person would
have put them, and seeing f you're right outguessingyourself on how well
they think. Look at their car, look at this, and, Hey, I was right!
The key to this metaphor is that even after the mind of the designer is "trained", it remain
very difficult to figure out and articulate the highly nonlinear relationships and rules
embodied in that black box; thus the statement above about not being able to figure out
the reason behind every response, for example.
As befits the highly open-ended and interpretive nature of the exercise they were
conducting, the members of the Color Studio, and our informant in particular, did not
have any qualms taking a piece of information-an observation or a response from a
participant-and reading into it a meaning or a conclusion (an admittedly provisional
one) that would not be supported from an analytical perspective. A typical example is the
way in which our informant was using the data from the gift certificate arrangement
described earlier. She explained that depending on which of the gift certificates a
223
participant requested, she and her colleagues could infer certain facts about that
participant:
We can track, eitherfashion, records, books, home, or, like,
organizational, Target, general... So then we can also kind of see, where
is their... Are they homeowners, are they...:?I don 't know, we got a lot of
Tower Records. 6
Needless to say, it would not be unheard of for someone who regularly spends an
important part of his or her income on home furnishings to decide to use a free gift
certificate to buy a music CD or a book, something they may not do often. To infer much
about a participant's interests based on which gift certificate they select seems tenuous.
Again, our informant was not concerned about this: she understood that she was not
conducting a statistical study, but simply using the different pieces of information
available as contextual elements for the interpretive process in which she was engaged.
The following quote reflects the questions she was struggling with:
I 'm going throughjust asking questions:Do they like this or did they
respond to the image? Do they hate it? Do they eat on the run? Do they
have a casual lifestyle? Do they not? You know. Some people like it, some
people hate it; maybe... Oh, this was actuallyquite interesting:young
women loved these things, these... loved them, every one of them. Older
people didn't care, and menhated it, but I'm not sure if they respondedto
the packaging, the gap, or like this indulgent kind of... I'm not sure. But
anyway, it leads me to think that they're very literal thinkers, this group:
what it is is what it is, in a weirdway; that they're reallyrespondingto it.
Finally, another aspect of the research exercise was described by our informant, which
highlights its interpretive nature. She explained that, after they were done sorting through
the responses and discussing them, she and her colleagues anticipated re-dividing the
chips among the four of them and re-organizing them using different classification
schemes, or different categories. One possibility she discussed: instead of using the
objects depicted in the images, they might define categories that reflect certain styles and
organize the chips that way.
62
Target is a chain of department stores that sell a broad range of products, including house wares, sporting
goods, clothes, and so on. (See www.target.com.)
224
Yeah, but then I think we'll break it apart again into like products. I was
thinking about doing things that... like maybe classic images or modern
things that... Define it that way and maybe do it the way we think and see
if they group into the same areasor if theyfall apart.
Our informant suggested a slew of questions and interpretations that would come out of
this process of taking a different look at the responses. Again, it was difficult to see how
these would be supported by the data and the research design, if one were to take an
analytical stance.
Drive-by shooting: A photo essay of random cars and their drivers
The exercise described in the previous section was conducted in the context of a specific
development project, aiming at developing a modern replacement for the Nissan Z car.
However, our informant described similar open-ended, interpretive research activities
within her studio, which were not connected to a specific project.
In this particular example, the Color Studio hired a professional photographer simply to
take pictures of people driving their cars. The photographer was asked to do that in two
different cities in the US. Following is our informant's description of this project:
... we worked with a photographer by the name of Andrew [B.] that
basicallyjust photographedpeople driving.He's an art photographer,he's
just fascinated by cultureand populationand changes and... We sent him
to a couple of different cities, and he drove around with his car, with his
camera mounted on the side basically, and he was very selective of what
he [shot.] It was everything and anything. And there is no science to this;
we didn't say, Go shoot BMW's, Go shoot this. Hejust shot everything.
And we have bindersand bindersofproof sheets that came back And I
put up all the proof sheets on the wall and we just stood back and--this is
really how really generalized--westood back and said, Wow,that's a lot of
red! I mean... That's a lot of teal! We see a lot of... I mean, the colors
popped off the page, and they were done like this; and then we said, Wow!
There'ssomethingjust really honestaboutstoppingsort of whatyou see
every day, and looking at it in a different way.
As was the case with the previous research exercise, our informant did not have specific
questions in mind that she was trying to answer by commissioning this project. She had a
63 Tower
Records is a chain of stores that sell music recordings and videos. (See www.towerrecords.com.)
225
general idea of photographing people driving their cars (she uses the term "generalized".)
And even though she describes her and her colleagues' surprise at the high incidence of
certain colors, this is an observation that emerged during the process. Clearly, their
objective could not have been to learn about the color distribution in the car fleets in the
two cities in question; such information is already tracked by a number of market
research companies, and NNA (the Nissan marketing organization at the time of the
interviews) in all likelihood already had that information. Furthermore, our informant
was unconcerned about any biases introduced by the photographer's
selection process. or
by the method used to look at the pictures (for example. while looking at a large number
of photographs placed side by side, one's attention is more likely to be attracted to the
brighter colored ones; ones depicting red cars for instance.) As she pointed out in the
above quote, "there is no science to this." Also, when pressed again on this issue:
}es, there's no--that's what I'm saying--there's absolutely no science to
this. He didn'tpurposelysort of go to sedans,or BMW's,or this or that; or
clean cars or dirty cars; or interestingpeople. I think the selectionprocess
obviously happened there...
Although it may seem from the above that the focus was on the cars in the photographs,
the idea was to take pictures of people driving. The Color Studio designers selected some
of the pictures for enlargement; in those, it was easy to see the drivers of the cars, what
they looked like, and what they were wearing. Again, the different designers took turns
sorting the pictures using different classification schemes. And in this case too, our
informant stressed the importance of the conversations that were taking place within the
Studio throughout this project, and the thoughts and ideas engendered by that process:
You know, maybe if we worked in this direction. as we are, that :sinspiring
to us and allows us to talk to each other and allows us to think more about
the product. And then ifsomeone else wants to come in and say. You're
.full of [expletive]! And, It doesn't mean anything! That s okay to me,
because that isn't why we did it.
Watching Oprah to stay in touch with the public
The discussion concerning mechanisms of external integration soon turned to questions
about ways in which our informant keeps in touch with trends and with what people are
226
doing. It should not be surprising by now that, here too, her approach is unusual. Her
description follows:
Exactly why I'm stuttering on this question, I'm not sure what it is. And I
think that each one of us is very, very different as designers. I know what I
do personally,and I think onceyou define it, maybeyou come up with a
dtferent way,you know... Takea sick day. I know it sounds really
ridiculous:I'm kind of answeringthis as I'm talking.On a sick day, I will
watch the shittiest TVyou canpossibly--Aaury Povich,Sally .lesse... I
will just go.from one to te other, one to the other... I love those shows. I
really.feelout of touch whei;I don't watchall that crap. And, maybe that
isn't an identifier:maybethat doesn't necessarilydraw it. I don't know
what it does, but somehow I like it. I like it. I like that. I like going to
places, public places wherethere's this mass of people, of all different
culturesand manyareas. Somethingthat isn't an identifier,not like a
museum or something that says you 're this type of person. Something
that 's so genericthat it 's-a parade or something-that there's no... I like
just watchingpeople, in general,masses ofpeople. I think it's great. I love
people.
As expected, she does not describe a situation where she relies on research reports and
consumer trends data. Rather, she engages in a set of activities which do not appear to
have, a priori. a defined set of well articulated deliverables. She does not suggest that the
purpose for watching a particular show is to identify popular apparel color combinations
in one specific demographic group. She does not even indicate that she observes what
people are wearing, for example, or what colors are used in the set or floats in the case of
the parade. Again, her focus is on observing people and what they are doing, not on
objects or things.
Another activity that was discussed in the same context is travel. Here too, her approach
is different from what other designers might do. Instead of visiting the large metropolitan
areas and the acknowledged of design such as Milan or Barcelona, she is more interested
in going to rather non-descript areas of the US, and to smaller, less touristy towns in
Italy, for example. More specifically, our informant described a trip she took with two of
her colleagues the previous year; she answered a question concerning any specific things
that she noticed, or ideas she got during that trip:
227
We ended up in P'ositano,a very small seaside community. [...] I can 't
think of anything right now. No, I can't, and maybe that s why I enjoyed it
so much, I'm not sure. I usuallyrememberwhere ideas and thingsstart
firom. And I don't think that I used that trip in that way. Maybe there 's
certain times that... I think I never not think about work: I don't think I
ever not think. But I don't remember looking.bfr that to be the solution to
something that I was... Maybe it was more like, a generalized thing as
opposed to a specitic thing I wanted to answer.
Also:
Yes. I think we need more travel. I don't think we travel enough in the US.
I don't think we travel enough to enough shitty parts of the country I want
to spend... It' a hard one to definebecause I haven't come up with a
strategy yet o/ what the company would get qfe I go through this process.
[...] Unless I come up with a plan. sometimes it' hard to get it.
This underscores the constraints that crop up in a business environment, when highly
interpretive, open-ended activities of the type discussed above come up against traditional
budgeting considerations, which are designed to deal with instrumental, means-ends
approaches. This was somewhat surprising considering the support afforded to the Color
Studio and the Design Context Lab by NDI's president, and his strong belief in, and
enthusiasm for, their approaches and methods.
228
Chapter VI: Case Study III
Nissan Technical Center
The Nissan Technical Center, or NTC, is Nissan's main design and engineering facility in
Japan. It is located in Atsugi, in Kanagawa Prefecture, about one hour outside of Tokyo
by train. As with the case at NDI, this study started out by focusing on the development
of the second generation Altima, but quickly evolved into a wide-ranging investigation of
product development practice at NTC. This report is divided into three sections. The first
covers the function of the Product Planning Department and the integrative role of the
"Shukan" or product development manager. The second section covers the area of design
(or styling.) Finally, the last section describes the "marketability" function, which falls
under the Product Experiment Department (a test and development department.)
Product Planning Department #2
The Product Planning Department (PPD) was described as the "center" that coordinates
and controls all product development activities. PPD personnel are in charge of all
decision-making as it relates to all proposed new Nissan models. This includes decisions
concerning vehicle size, performance, features, engine size and configuration,
transmission choices, and all other specifications. PPD personnel are also responsible for
deciding on the investment level and profitability targets for these new models. The PPD
is organized into two directorates, #1 or "Luxury" and #2 or "Global", each of which has
its own design and development groups, such as body design, testing, and so on. The
Altima falls under the "Global" directorate, along with other models that are popular in
different world markets, and which therefore need to be developed "on a worldwide
basis." These include the Maxima and Sentra sedans, in addition to the Altima.
229
Generally speaking, there is a Shukan, or product manager, for every vehicle model,
reporting to one of the two PPD directors. Our two informants in this part of the study
were the previous Altima Shukan, who had recently been promoted to Deputy Manager
of the Product Design Administration (discussed later in this section). and his successor.
The Shukan is typically an engineer promoted from one of the development groups. such
as testing, body engineering, powertrain, etc. A few have come from the sales
department. and one came from exterior design. The senior informant indicated that in
the future, Shukans would most likely come tfrom engineering because. as he put it, "the
characteristics of a new vehicle model are mostly influenced by the design engineering
circle" (referring to one of several finctional circles he had sketched.)
In addition to the PPD. there is a separate office, the Product and Marketing Strategy
Office, or PMSO, which is in charge of deciding the positioning of each Nissan model.
The PMSO gives the PPD a specification range for each model, and the PPD then decides
on a particular point within that range.
Our informants focused on describing the development process, in particular the process
through which the Shukan determines the vehicle features and specification targets
mentioned above. At the launch of a new project, a "role-studying" committee is
assembled to decide on the role of the proposed new model. A Shukan is nominated for
that model, if one is not already in place. This committee is multidisciplinary, with
participants from the PPD, PMSO, engineers who are in charge of vehicle packaging
(platform and body engineering), and representatives from sales, who are in charge of
putting together the business case for that model. The outcome of this first phase is a
brief with high-level specifications for product position, price, size, engine, cost, and so
on.
In a second phase, a development study request is issued to the various engineering
groups-design engineering, production engineering, manufacturing-to conduct
preliminary studies based on the initial development brief. The results of this study are
fed back to the PPD, which then generates a first draft of the product specifications.
The draft product specifications are used as the basis for the start of the development
process proper. In this third phase, each department undertakes more refined engineering
230
studies. A significant difference at this stage is that now, for the first time, a styling study
is also undertaken as part of the development work. The information that results from
these studies is again fed back to the PPD (this time including a design proposal), and the
draft product specifications are refined. This last phase may involve a few iterations in
order to refine the design proposal.
For the purposes of this dissertation, the details of the process itself are not critical. Of
greater interest is the fact that our informants chose to focus on the process itself, not its
content or particulars as they played out in the case of the Altima project. The rigidly
structured, pre-planned information feedback and decision points received particular
emphasis.
Equally interesting are the sources and types of market information that our informants
stressed. In the case of the Altima, they described a set of traditional approaches to
obtaining customer information, including the use of sales and demographic data, visits to
US Nissan dealers, and user group interviews. Unlike NDI's approach to research where
the Design Context Lab personnel and the designers themselves were directly involved in
conversations with customers, the Shukan described a particular focus group where he
stood behind a one-way mirror as Altima owners were interviewed by professional
facilitators.
Our informants reported that PPD does not conduct any activities aimed at learning about
trends in the marketplace, outside of projects tied to the development of specific vehicle
models. That is not the case with the PMSO, which conducts regular market surveys.
However, here too, the approach used is an analytical one, and the PMSO's role appears
to a rather mechanistic one: one of the two leading market research firms in Japan are
usually retained for the job; the focus is on collecting demographic data-age, gender,
income level, among others-as well as data on customer satisfaction for the various
Nissan models (and possibly their direct competitors;) the resulting reports from the
research firms are then circulated among the relevant PPD personnel.
Although not related to product planning, the new role of the outgoing Altima Shukan
was also discussed. He described his new position of Deputy Manager of the Product
Development Administration (PDA) as a "conventional position". He is engaged in
231
identifying commonalities between the various groups involved in product development.
This includes those that fall under the two directorates described above, with their design
and testing groups, as well as other development related groups within Nissan. The
purpose of this work is to reduce overlap and redundancy between these groups, and to
rationalize and shorten their processes. In other words, the task of the Product
Development Administration is an analytical one, aimed at streamlining the development
process and making it more efficient.
The PDA also serves as the contact point for the control and dissemination of product
development information. Again, this appears to be a largely mechanistic activity. All
design and engineering releases go through the PDA. and these typically follow preagreed schedules. For example, product engineering may be required to make certain
decisions or to turn in a certain set of specifications to the Design Administration by a
given date. If this is not done on time, the PDA intervenes to try to identify the problem
and its cause, and issues a warning to the Shukan to "hurry up." Although the PDA
oversees the schedules of all development projects and tracks their progress, it does not
decide the schedules themselves. Its task is a structural one that involves information
gathering, dissemination and control, but that is not concerned much with the content or
meaning of that information. From the PDA's perspective, the schedules and rules are
pre-agreed or pre-arranged externally.
Exterior Design Department #2
A second part of this study dealt with the development process from the point of view of
exterior (and interior) design. Our informant was the "Design Producer" for the Altima
project. He is a senior manager in Exterior Design Dept. #2, which is responsible for the
"Global" cars. In that position, he works with the Shukan to "negotiate the appearance,
specifications, and features of the car," and he oversees all the design teams working on
it, including the exterior, interior, and color teams at NTC in Japan. and, indirectly, the
exterior design and the color teams at NDI. (NDI did not have a team available to
participate in the Altima's interior design.)
232
Work on a new model starts informally before the official Design Declaration point, with
the "advanced stage models." These are sketches intended to serve as an "evaluation of
the current product," and to provide a hypothesis or hypotheses that "the desightlrs want
to check with the market." When the PPD officially announces the start of the
development project, the Design Proposal stage begins. This stage typically involves
between twelve to fifteen sketches, followed by two to three quarter-scale models. One or
two of these are then selected by the Design Strategy Committee to be developed into
full-scale models. This leads to the Model Direction Selection decision point, which our
informant described as "the biggest event" as far as exterior design is concerned. The
full-scale model that is chosen is then developed into a detailed production clay model. It
is refined until the Model Freeze point is reached.
The various decision points in going from sketches to quarter-scale clay models to a fullscale clay, and the positions of the decision makers involved at each step, are very similar
to those found at Chrysler (please refer to Chapter IV) and the rest of the industry. Sketch
and quarter-scale model selection are mainly done by NTC designers along with key
executives in charge product development-mostly personnel from PPD. Model
Direction Selection involves executives at the Vice President level, specifically the VP of
Sales and the VP of Engineering, as well as personnel from the relevant marketing and
manufacturing subsidiaries, Nissan North America and Nissan Motor Company (both in
the US) in the case of the Altima. One significant difference from Chrysler is the
importance Nissan attaches to the voice of the customer. Findings from consumer surveys
and consumer clinics play an important role in these decisions. Our informant had an
interesting perspective on this aspect and its causes:
Nissanrelies on clinicsfor model decision.Because of the late 80s
successes, the company became less diverse, more stiff My personal
opinion is that this is due to the fact that the company did not have enough
experience with success. It was ourfirst success, and as a result we
became cautious; we did not want to lose that success. So we told each
other, " We can 'tfail now. " So we started relying on more market
analysis.
233
Not only is this cautious attitude in sharp contrast with Chrysler's risk taking mentality;
its "stiffness" is also at odds with NDI's creative, unorthodox, even playful approaches to
learning about the customer.
Our informant also blamed this cautious, analytical approach, and the over-reliance on
customer input in the design process for the difficulty his company was facing in defining
itself and the character and look of its products, a problem that several Japanese
manufacturers share:
Ifyou look at the total informationthat goes into producinga product,
[...] the currentpattern of Japanese companiesis 20% creationand 80%
market information, which is the same to everybody-Nissan, Toyola.
Maybe Honda is different. So the differentiation is only in that 20%. We
need to go to 60-40 in order to get our own identity.
He traced the problem to the recent history of Japan's automobile companies. According
to our informants, starting in 1945, Japanese companies have targeted the US and
European markets. Their goal therefore was to catch up to American and European
designs. Although he admitted that in certain areas they were still somewhat behind, he
felt that, in general, Japanese designers had caught up. However, he believes that they
had done so at
high price:
For more than 50 years, we have been chasing and copying their way of
doing, so it has become our nature [to follow and copy.] Even now, when
we are told that we can go anywhere [with a design], we can 't. So we 're
having a very tough time deciding where we, as Nissan, want to go.
Aside from copying foreign designs, our informant related another approach to product
positioning, based on a metaphor form golf, which he believes has been commonly used
by Japanese companies. His description reveals an analytical process of averaging or
interpolation. And again, he blames the resulting thinking for his company's difficulty in
defining itself:
It 's the "OB " strategy,from golf terminology. This has been Japan 's
strategy in PD since 1945. Ifyou don 't know where to go, you look at your
competitors and see where they are, and you aim in between them. There
is one here, and the other is there, etc. I will hit the middle. The question is
how to define myself? Up to now, it has been like this: first, define the
234
competitors; second, define ourselves. But now, we need to define
ourselvesfirst!
Throughout the discussion, attempts were made to surface non-analytical practices at
NTC. Unlike the case at NDI, the studio designers at NTC do not undertake any non-
automotive design projects. Furthermore, most of their work is tied to regular production
models. Our informant reported one set of exceptions, dating back to the late 1980's and
early 1990's. These are the so-called "Pike" cars, the BE-1, Escargot, and Figaro. The
latter two are a micro car and micro van inspired by the original Citroen 2CV. They were
the result of some work that a few young designers did after hours, on their own time.
The Be-1, a playful mini car combining retro and modem elements, came about after a
few designers spent some time with outside consultants, discussing trends and lifestyles,
and doing a lot of "town watching." These young designers then translated the ideas they
had following these activities into design proposals and show cars, which led to the Be-I
going into limited production.
These activities appear to have been limited to that particular time period, and they were
probably connected to the company's economic success in the late 1980s. There is no
evidence that similar activities were ongoing at the time of this study.
Our informant stressed the importance of helping designers stay in touch with popular
trends, in general as well as in the automotive field specifically. He acknowledged that
NTC's location, which he described as "quite remote from downtown Tokyo," is not
particularly conducive to that. One answer is to "pick certain young people and middle
managers and send them to refresh or recharge their good designs." This is done when
there is a lull in the workload, outside of project peak times. There is no official rule for
who gets chosen, but "the more senior managers are always watching to see who is
exhausted and who needs refreshing."
Another solution relayed by our informant is what he called "mini Detroit" or "mini Paris
shows", referring to two of the more important industry shows, which are held in these
cities. He described these events as follows:
Recently, we have been doing what we call mini-Detroit or mini-Paris
shows. What we do is, we get ten cars from Europe orfrom Detroit, and
235
we display them in the hall, so even the young designers and engineers get
to see all thesemodels that are not availablein Japan. If l had enough
money, I would send all the designers to Italy and Europe, to the market in
question. Of course, to see them in their environment would be ideal, but
still, this is better than lookingat pictures.
Finally, our informant described an interesting system, the Senior Designer system,
which he referred to as "unique". Senior designers, of whom there are ten at Nissan, are
different in that their activities are not confined to Atsugi. They have an expense
allowance, something that file regular studio designers do not have. One senior designer
uses his allowance to keep a studio in Tokyo; others have used their allowance to go to
Italy to work on particular projects, either to do sketch work or quarter-scale model work
(a senior designer can select a modeler to accompany him on such trips.)
In general, our informant gave the impression of wanting to create space for more
interpretive activities to take place at NTC. and wanting to encourage more activities that
would immerse his designers in the environments in which their customers live. This is
apparent in his comments concerning NTC's distance from Tokyo. for example, and
concerning the suboptimal quality of the mini auto shows. But this desire appeared
hampered by a rigid, analytical, cautious corporate culture, which results in a set of halfmeasures. Studio designers are (temporarily) given more freedom in order to solve a
burnout problem, not to get them immersed in a different environment proactively in the
hope of stimulating their interpretive thinking. The Senior Designer program is another
telling example: a designer has to reach a certain level of seniority and accomplishment
before he is allowed the freedom that may very well contribute to that high level of
achievement in the first place.
Product Experiment Department #2
The third part of the study dealt with what Nissan calls "Marketability Testing", a
competitive testing and evaluation function. This function falls under the Product
Experiment Department, another name for product testing and development. Our
informant was the person in charge of the marketability testing area within Product
Experiment Department #2, the department responsible for the Altima.
236
Our informant characterized Marketability Testing as the interface between the Product
Planning Department and the Product Experiment Department. Whereas the PPD and the
Marketing Department interact directly to determine the positioning and features of the
proposed product, when it comes to vehicle performance, the Marketability Testing group
plays a crucial role.
Marketability Testing (MT) as described by our informant is a data driven activity,
similar in many ways to quantitative market research. The MT group develops
questionnaires and conducts surveys about different aspects of performance, targeting
both current Nissan customers as well as owners of competitors' cars. Occasionally,
market research companies are hired to conduct the research. Based on the data collected,
they set certain performance targets and they propose them to the Shukan. Later on, after
goals are set and prototype vehicles are built, MT gets involved in testing them and
confirming to the Shukan whether the original goals have been met.
Our informant explained his role in decision making thus:
Whena decisionis made, there are many considerationsthat enter into it:
there is time, cost, and quality considerations. But sometimes, the
decisionsare companyoriented instead of beingcustomeroriented In
that case, I give the viewpoint of the customer. I use competitor [product]
data, or use data from our own research, to try and attach more
importance to the customer 's side in that decision.
Sometimes, playing that role involves setting up and coordinating specific tests to resolve
conflicts. As an example, our informant gave a case where different members of the
development team were in disagreement concerning the noise level of a model under
development. The MT group set up a test and invited a group of customers to try out and
comment on the different setups. The findings were then presented to the Shukan who
made the final decision.
Although the voice of the customer is generally used as the basis for resolving design
conflicts, our informant explained that this is not always the case. On many occasions,
decisions are made that run counter to the customer's preference. But even then, the
reasoning is analytical, with the decision based on a straightforward cost-benefit
calculation: following the voice of the customer would simply be cost prohibitive. There
237
was no indication that such a decision might result from a different interpretation of what
the customer really wants; nor, unlike the case of Chrysler, that it might result from the
designers taking a risk based on their belief that they have a better understanding of what
the customer will want in the future than the customer himself.
In general, our informant described an approach to decision making that was very
structured and highly planned, where goals and targets were set at the very beginning of
the project, mainly based on survey data. Goals are organized by importance,
distinguishing between "must haves" and "wants", for example. Analytical criteria are
developed early on in the process, for making the allowable tradeoffs or concessions as
conflicts emerge.
Certain aspects of the approach used by the MT group to analyze the voice of the
customer do appear to have an interpretive dimension. The results of the surveys
described earlier are organized using a tabulation scheme similar to the House of Quality
(Hauser et al., 1988). For each product attribute on which the customer is surveyed, a
numerical value is entered in the table, reflecting the score of the current product. If data
is available for competing products, those are entered in the table as well. Values are also
entered for the importance the customer attaches to that attribute, and for his or her level
of satisfaction with that aspect of the product. An interesting interpretive aspect of this
table is that it is divided into different segments, corresponding to what our informant
called different "scenes." These represent different contexts of use, such as highway
driving, driving on a winding road, around a parking lot, and even washing the car. The
table also includes information about the frequency of occurrence of these different
scenes under normal conditions of use. Furthermore, the table includes a set of entries
that indicate how the customer's voice should be "interpreted". Our informant explained
that different customers might mean different things even when they use the same terms.
For example, to certain drivers "sporty" means crisp handling on winding roads. whereas
to others it means good straight-line acceleration. Similarly, a sedan driver expects a
lower level of interior noise before he or she would call that car "quiet", compared to a
sports car driver. In order to account for such differences, the table provides for entries
about "how to interpret" the findings regarding certain attributes. In both cases, however,
238
the interpretive elements are reduced to numerical scaling factors, which are used to
adjust the numerical scores entered in the table for the various attributes. Clearly, this is
very different from the open-ended interpretive practices found at NDI, such as the
approach used by the Design Context Lab to understand the context of use of their future
mini SUV (please refer to Chapter V.) The concrete examples selected by our informant
to illustrate the use of these data in differentiating the Altima from one of its competitors
confirm the basically analytical nature of this approach. He focused on purely technical
aspects of the product, namely, the size of the cup holder and the size and output of the
motor. His description follows:
Concrete example? The Accord is very strong: that is difficult. For
instance,just before the 96 model year, the Accord could not accept a
mug[-size] cup, so that s why we added a cup holder to the new Altima
that could hold a mug. Unfjrtunately,they madea minorchange and
revised their cup holder size. [...] They have a 2.2L displacement engine,
we have a 2. 4L; the advantage is really good [response at] freeway speed
They have good engine technology, and really good pickup; we don 't
unjbrtunately. But on the other hand, we have a 200cc advantage, and this
is alsopretty good atfreeway speed
239
240
Chapter VII: Summary and Conclusions
This chapter summarizes the findings of this dissertation, both theoretical and empirical.
It includes recommendations for product development researchers and practitioners, and
suggests several directions for future research.
Findings
The findings from this research are presented in two sections. In the first one, I
summarize the theoretical developments that emerged from the empirical evidence, and
in the second, I summarize the empirical findings themselves.
Theory development
The theoretical framework developed in this work takes the form of a typology, or
taxonomy, which attempts to go beyond the generally accepted dichotomy between
positivist and interpretivist views of product development. The taxonomy draws upon
concepts and ideas from linguistics and from the philosophy of language. In a first step,
the distinction within the field of linguistics between the structural sub fields (specifically
syntax and semantics) and the functional sub field of Pragmatics is used to sharpen the
difference between analytical/structural practices on the one hand, and interpretive
practices on the other. In a second step, two views of interpretation, one grounded in
linguistics, and specifically Pragmatics, the other grounded in the philosophical
hermeneutics of Heidegger and Gadamer, are used to expand the interpretive category
into two, which I refer to as pragmatic interpretation and hermeneutic interpretation
respectively, after Dascal (1989).
The use of the different aspects of linguistics and language as a model for structuring the
theoretical framework is felicitous, given the importance of communication and
241
interpretation in design and product development (Katz, 1982; Krippendorff, 1989;
Dougherty, 1990; Griffin & Hauser, 1990; Moenaert et al., 1990; Dougherty, 1992a;
Brown et al., 1995). This particular use of language in the area of design is very different
from the atomic language models of design. as developed by Stiny. Mitchell and Wallace,,
where structural linguistics models of our natural language were used as a pattern for
developing shape and style grammars, that is, formal, logic based models of particular
design worlds (Stiny et al., 1978; Stiny, 1980; Mitchell, 1990; Wallace, 1991a). This
work is closer to the concept of product semantics advanced by Krippendorff (1989).
Although reasonably parsimonious, the taxonomy presented in this research is widely
applicable and can be deployed in a number of areas. I used it in describing a designer's
thinking process-the
internal conversations she was having with herself and her work
input and product-as
she reflected on what she and her colleagues were doing (Schon,
1983). 1 also used the taxonomy in describing the interactions between the different
members of a design team, specifically those from different functional areas, in order to
discuss various approaches to internal integration. In the context of external integration, I
used it to discuss the relationship between a company and one of its suppliers. I also used
it extensively in analyzing the approaches used by the various product development
practitioners to learn about and develop a better understanding of their customers.
Generally speaking, the typology should be applicable in most situations that involve
people communicating, whether they are exchanging well-defined bits of information, or
whether they are engaged in a more open-ended process of coming to understanding.
The taxonomy was used briefly to analyze the evolution of two methods that have been
proposed in the last few years for improving the organization of product development
projects, namely, the stage-gate system (Cooper. 1990; O'Connor. 1994), and the design
structure matrix or DSM (Eppinger. 1991; Eppinger et al.. 1994). An interesting finding
from that exercise was an evolutionary path shared by both methods, which comes across
clearly in the respective literatures. Both methods were initially presented as analytical.
structural models of the product development process. with the content of the models
treated as external inputs to the model and assumed to be unambiguous. In the case of the
stage-gate system, objectives and other inputs were assumed known at the beginning of
242
the process, and decision criteria were assumed clearly and unequivocally defined at the
beginning of each stage. In the case of the DSM, the tasks and their informational
relationships were assumed fully defined, for example from a physical model of the
product. Later on, as more experience was gained with applying these methods in
practice, they evolved into approaches that have important pragmatic interpretive
dimensions. An updated version of the stage-gate system made room for overlapping
stages and for fuzzy and conditional decision points with emergent decision criteria
(Cooper, 1994). In the case of the DSM, one focus of research became the interview
process for obtaining meaningful or "credible" information about the relationship
between activities, in other words, the interpretive modeling process used to develop the
DSM (Dong, 1999).
From a pedagogical point of view, one of the goals of this work was to enable
practitioners and students of design and product development to understand the
advantages and disadvantages, as well as the limitations and range of applicability of the
countless tools and methods proposed to them. In particular, I wanted to make this work
accessible to engineering students of various backgrounds. To that end, the theoretical
framework was presented as a self-contained (albeit lengthy) chapter. It does not assume
much (if any) prior knowledge or familiarity with the concepts used to develop the
taxonomy, neither linguistics nor hermeneutics.
Empirical findings
In considering the empirical findings presented here, the reader is invited to keep in mind
the theory building nature of this research. These findings should therefore be considered
as hypotheses to be tested in future research, as opposed to generalized findings
supported by statistical replication (Eisenhardt, 1989).
The first notable observation is that each of the three organizations studied in this
dissertation has a propensity towards relying on one of the three classes of product
development approaches developed in this thesis. Chrysler is essentially a pragmatic
interpretative organization. It is not likely to take customer data at face value and accept
it without questioning it or without putting its own twist on it; but it does so in a rather
directed, focused fashion. At the same time, it does not come across as a company likely
243
to indulge in open-ended pursuits. Nissan's NDI, on the other hand, represents the
consummate hermeneutic interpretive environment, where designers are given the
freedom to engage in highly open-ended projects, the outcome of which is difficult to
predict or explain beforehand. Finally, Nissan's Technical Center in Japan has the
hallmarks of a mostly analytical organization, at least judging from the relatively limited
case study reported in Chapter VI. The focus there is on obtaining very specific survey
data from consumers, and following the data to the extent possible, given cost constraints.
The next finding is that practices utilizing hermeneutic interpretation were more common
in situations involving styling and color, viz., the NDI Color Studio. One explanation
points to the nature of shape and color, and the difficulty of describing them verbally in a
precise and objective manner, with one person's description often unlikely to match
another's. This difficulty in quickly achieving intersubjectivity on matters involving
form, color, and texture makes a hermeneutic interpretive approach better suited for
carrying out a conversation about them, as it naturally allows the vocabulary and
categories to emerge during the conversation. Another explanation is that thes, are areas
where the designer enjoys the greatest degree of freedom, and where there is the most
room for creativity, which makes open-ended, hermeneutic approaches particularly
applicable. Designing a new automotive shape or a new color palette are activities that
feel least like traditional engineering problem solving. Even if one were to cast these
activities in terms of problems to be solved, one would be hard pressed to provide criteria
for determining when the problem would be considered solved, that is, what would
constitute an acceptable solution.
A third observation is the predominance of hermeneutic interpretation in those situations
that involve external integration, such as the unusual approach used by NDI's Color
Studio and Design Context Lab to understand the customer. As with the previous finding,
two explanations. similar to the ones discussed above, present themselves. One
explanation is the language barrier: if designers and developers already have difficulties
discussing shape and color among themselves, the situation would be worse if the
customer, who is removed from their world and unfamiliar with their specialized
language, were invited to join in such a discussion. This would explain why, at Chrysler
244
for example, only trained designers are invited to comment on proposed designs (please
refer to Chapter IV.) Another explanation is the open-ended mandate that the developers
were pursuing in these particular cases (the future Nissan mini SUV and sports car), and
the wide ranging exploration they were conducting in the hope of understanding new
consumer trends, and of discovering previously unknown or unarticulated customer needs
and wants. In particular, they were consciously questioning their preexisting
understandings ("Let's check our own intuitions") around these product categories, in
view of the then recent breakdown in the sports car market.
Conversely, situations of internal integration were more likely to call for pragmatic
interpretation, as they were closer to situations involving relatively directed problem
solving. One example is the work done by NDI's Design Context Lab in developing the
"outer bounds" document for an environmentally friendly vehicle, as described in
Chapter V. In that case, a product description was needed, that would bridge the gap
between the marketing organization and the designers at NDI. Here too, the crux of the
problem was a gap in language between the two groups; however, the situation involved
a well-defined problem, resulting in an activity that was targeted. with a clear objective
and an ex-ante understanding of what a successful outcome would entail. Another
example is Chrysler's decision to include a 4 th door in its Minivan design. In that
example, the interpretive challenge was less one of understanding the voice of the
customers, who were clearly and unambiguously opposed to the idea, as one of bringing
the team to a consensus regarding their decision to go against the customers' wishes.
The fourth observation is the important role of trust in enabling the use of interpretive
approaches in design and product development. This is not surprising given the
speculative nature of interpretive activities. Trust among team members, as well as
between the development team and upper management. was mentioned on several
occasions by our informants at Chrysler. At NDI, although the issue of trust did not come
up much, it was obvious from the interview with NDI's president that he had managed to
create a non-politicized, open, even collegial culture within the organization. Recall for
example that he encouraged the designers regularly to rummage through each other's
desks. It is also clear that he was able to shield the designers from the influences of the
245
parent company, as well as create a culture that encouraged and supported their highly
interpretive activities.
Limitations
As with most research, this work suffers from several limitations. One set of limitations
stems from the research approach used. Eisenhardt ascribes two potential pitfalls to
theory building. The first is that the outcome might be an overly complex theory due to
the intensive reliance on empirical evidence. The second is that the resulting theory may
be "narrow and idiosyncratic", tied to specific phenomena due to the bottom-up nature of
theory building (Eisenhardt, 1989, p. 547). I have attempted to avoid these pitfalls by
limiting the theory to a relatively simple typology, one more likely to be broadly
applicable. I have used this typology to organize, explain, and discuss the findings,
without extracting specific organizational guidelines or recommendations. For instance,
the findings at NDI are presented solely as examples of interpretive practices, without the
suggestion that the way NDI is organized, or the particular practices used there, are the
only appropriate ones, or that they ought to be duplicated by companies facing similar
situations. Similarly, although I propose that a company ought to consider separating its
analytical activities from its interpretive ones as it structures its supplier relations, I do
not suggest that the model to follow is necessarily the one observed at Chrysler and Lear.
To the contrary, it may be that a limitation of this work is that it leans too far towards
generality, because what is proposed is a descriptive typology, rather than a set of directly
actionable prescriptions or methodologies. Practitioners can learn from the theory how to
distinguish the different types of situations they may face, and to be aware of the
limitations of the tools and approaches presented to them. They can also learn from
studying the rich examples presented in the case studies, about how certain organizations
dealt with different design and development situations. However, modifying the
approaches presented here or developing different ones to suit their particular needs still
involves a creative leap, an extra step that is context dependent. I would argue that this
creative step may be more challenging than mechanistically applying an analytical
methodology such as the house of quality or the design structure matrix.
246
Finally, certain limitations can be traced to the particular industry chosen for the case
studies, namely the automobile industry, and to the significant focus on the industrial
design (or styling) aspect of automobile design. In Chapter II, I argued, following Clark
and Fujimoto (1991 ), that the internal and external complexity of the automobile makes it
a good choice for studying design and product development in general. Still, automobile
design is principally evolutionary in nature. Although recent years have seen the
emergence of new popular product categories such as sport-utility vehicles and new
hybrid designs such as the Audi Allroad (Knoll, 2001) and the Volvo V70 Cross Country
(Passell, 1998), the basic architecture of the product has not changed in decades.
Similarly, the basic technology, namely, the internal combustion engine and the welded
steel structure, have remained static. As a result, there is a greater level of specialization
and compartmentalization within the automobile industry between industrial design
(styling) and the engineering design activities (Tovey, 1992). Case in point, Nissan's NDI
and Chrysler's Design Office are separate corporate entities. The situation is very
different in the typical industrial design firm, such as IDEO or Design Continuum. There,
engineers and industrial designers typically work closely side by side (Rosenberg, 1997).
Still, this limitation is mitigated by the fact that the industry had already started moving
to closer functional integration by the time the empirical research was conducted, as we
saw in the case of Chrysler's reliance on multidisciplinary platform team.
Implications
For researchers
This dissertation represents an example of a multiparadigmatic approach to the study of
product design and development. The findings, both theoretical and empirical, underline
the power of such an approach, and the many ways in which using a single paradigm
would be limiting. The comparisons between the different sites and their different
approaches to product development would have been impossible to capture and contrast
without the breadth of perspective provided by this theoretical framework.
The approach used here is not the mode-switching or multiple-lenses approach suggested
by some researchers. In that approach, the situation of interest is studied first from one
247
self-contained perspective or paradigm and, in a subsequent step, studied from another
paradigm. Later, the findings and conclusions obtained from each sub studies are
compared (Gioia et al., 1990; Dorst et al., 1996). By definition, paradigms are
incommensurate; each provides its own proprietary concepts and vocabulary, which do
not accept any intermingling. However, in the real world, successful product design and
development involves different types of approaches, each of which has an affinity to a
particular research paradigm. This thesis suggests a framework for bridging the gaps
between the concepts that the different paradigms make available to the researcher.
The framework presented in this dissertation could be useful to researchers engaged in
developing tools, methodologies and process models for design and product
development. It provides the vocabulary and categories that would help in understanding
the range of applicability and the limitations of such methodologies, following the
examples of the stage-gate process model and the Design Structure Matrix presented at
the end of Chapter III. It also suggests ways in which these models could be extended.
For example, the framework would give a more solid theoretical grounding to models
such as Coopers "third-generation"
stage gate process model, in which he proposes the
use of "fuzzy" gates and overlapping stages in order to deal with some of the openendedness inherent in product development practice (Cooper, 1994).
For PD managersand practitioners
The case studies in this dissertation show that it is possible to find, within the same
industry, the entire spectrum of design and development practices represented by the
typology developed here. This is the case even in an industry as well established or
mtiture as the automobile industry, where new products entail little innovation or novelty.
In fact, the two organizations that provided the examples reflecting the extremes of that
typology, the Nissan Technical Center and Nissan Design International. are part of the
same corporate entity.
However, the findings also suggest that using one or another type of approach to product
development is not simply a matter of making a simple choice at the appropriate time.
Different organization are simply better attuned to, and adept at, one of the different types
of practices. This reflects a number of factors, from the background and experience of the
248
employees, to the organization's approach to budgeting, to the value system instilled by
its senior managers. The impression given by Chrysler and NDI is that these
organizations are coherent and internally consistent. The attitudes of the informants, their
approach to their work, and the character of the products they produce are consistent with
each other and with the image of the organization and its senior executives. Similarly, the
research approaches used at the NDI Color Studio were not only supported by NDI's
president; they also reflected his views on the nature of design as an activity and those of
the people he had hired over the years. It is difficult to imagine the Chrysler informants
operating comfortably in an environment like NDI. It is even more difficult to imagine
the NDI designers working happily in the highly structured, analytical environment of the
Nissan Technical Center in Japan, with which, ironically, they interact regularly.
This research highlights several challenges to managers of product development
organizations. The first one is the need to understand the source of the organization's
current or future comparative advantage, using the typology developed here. Does the
organization derive its market advantage primarily from its analytical activities, or from
interpretive ones? And does it see this changing in the future? Clearly, in certain product
areas, the automobile with its complex interfaces being one example, the top tier
producers may not have the luxury of focusing on one type or the other. This drove
Nissan to found NDI in the first place, even though the president of Nissan at the time
may not have seen the distinction between Nissan's work in Japan and the work to be
done at NDI in terms of analytical versus interpretive design work. The fact that several
companies in the industry have followed suit by setting up design organizations in
California and in Europe (Barron, 1998) underlies the realization by these companies that
interpretive design and development activities are an important source of market
advantage in the current environment.
Other companies, lacking the ability or experience to carry out interpretive design and
development activities in-house, often choose to contract that type of work to specialist
houses (Tagliabue. 2001). This is especially true of certain automotive companies that
had previously focused on being low-cost producers (a mostly analytical challenge), and
later chose to compete more directly with the top-tier manufacturers. One such example
249
is Isuzu, a second tier Japanese company, which relied on Giorgetto Giugiaro's Itladesign
for design (interior and exterior styling), and on Lotus Engineering in England for chassis
tuning for its cars of the late 1980s and early 19 9 s.64
65 (In 1995, Isuzu decided to stop
0
producing sedans and to focus on trucks and SUVs.) A more recent example is Korea's
Daewoo with its Leganza sedan, introduced to the US market in 1998. Here again,
Italdesign was retained to help with the styling (Siano, 1998).
In a sense, the typology of product design and development presented in this dissertation
provides managers with another perspective on the core competence and sources of
comparative advantage of their organization. As we saw in the case of Chrysler and its
relationship to its interior suppliers, this perspective can be more insightful than the
traditional views where core competence is tied to a specific technology, subsystem, or to
generic integration capability (Prahalad & Hamel, 1990; Stalk. Evans, & Shulman, 1992).
In the Chrysler case, the insight came from drawing a distinction between an analytical or
structural integration role, which Chrysler was comfortable turning over to its supplier,
and a more interpretive integration function that it viewed as a source of market
advantage and chose to keep in-house.
The typology presented here also provides managers with a useful framework for
considering how to locate the different groups involved in product development. Several
studies have highlighted the importance of collocated teams as a success factor in product
development projects. This dissertation suggests that collocation may be particularly
important for interpretive design and development activities. And as companies
increasingly undertake product development projects that involve geographically
dispersed groups, decisions concerning the division of labor among these groups and the
choice of IT collaboration tools may benefit from an understanding of the analytical and
interpretive dimensions of the work in question.66
i
For example, the Isuzu Impulse, which was based on Giugiaro's Asso di Fiori prototype, shown at the
1979 Geneva Auto Show.
65
As explained in an earlier chapter, chassis tuning is considered more art then science, and generally
requires experienced vehicle test and development specialists. It is closer to an interpretive activity than
it is to solving a mathematical optimization problem.
66Large multinationals, such as Ford and General Motors, are trying to maximize the utilization of their
development resources that are dispersed around the globe. These efforts have gained momentum for a
250
Regardless of how labor is divided and workgroups are organized, the challenge of
integrating those involved in interpretive activities and those doing analytical work
remains. The case studies in this dissertation provide a different perspective on what has
been traditionally seen as the source of many integration difficulties, namely, the usual
schism between analytical engineering thinking and the more "intuitive" or "non
rational" thinking of designers and stylists. They show engineers, marketers and
designers all relying on a range of approaches, from the highly analytical to the highly
interpretive, depending on the situation and the environment, with some favoring one
type over another. The Design Context Lab at Nissan Design International provided some
interesting examples of how one company approached the issue of integrating between
the interpretive thinkers at NDI, and the more analytical marketers at Nissan North
America. An important insight that comes form these examples is that the integration task
itself is an interpretive activity, one that is best entrusted to people comfortable with that
approach.
At a more practical level, this research highlights another challenge that product
development managers must address when the organization undertakes both analytical
and interpretive activities. The issue in question is how to budget for highly open-ended
(or hermeneutic) interpretive work, when the outcome of that work is not clear even to
the people undertaking it. As our informant at the NDI Color Studio indicated towards
the end of the interview, the very nature of hermeneutic work makes it difficult to explain
and justify, especially in comparison to more analytical activities with their well-defined
tasks and goals.
Suggestions for Future Research
One set of future research projects immediately suggests itself by virtue of the research
design used in this dissertation. The obvious next step would be to take the typology and
concepts developed in this theory building exercise, and use them in a larger sample of
companies to test the observations and hypotheses advanced in this dissertation. Some of
the findings worth studying include: the degree of reliance of various companies on the
variety of reasons, including advances in information technology, and a reduction in certain types of
product differences across geographical markets (partly due to more uniform regulations.)
251
different types of design practices; the role of upper management and the corporate
culture in enabling or hindering the reliance particular types of practices; the
preponderance of hermeneutic interpretation in areas of aesthetic design or styling; the
relationship between the type of product being designed and the type of approach used
(for example, comparing a utilitarian van versus a sports car intended to evoke an
emotional response in the customer.)
One of those areas of research that is worth singling out is the relationship between
interpretive practices and creativity and the development of innovative designs and
solutions. How do the concepts advanced here relate to the rich body of research on
creativity, from the highly analytical (Navinchandra, 1991; Finke, Ward, & Smith, 1992;
Boden, 1994) to the less so (Gardner, 1994; Csikszentmihalyi,
1996)?
A second avenue of research would be to study the applicability and relevance of the
theoretical framework and its typology in industries other than the automotive. Although
the larger project within which this research was conducted studied a broad range of
industries, from apparel to medical devices (Piore et al.. 1995, 1997b, a). that work was
conducted using an earlier version of the framework that did not distinguish between
different interpretive approaches (Piore et al.. 1994). It may be fruitful to revisit the
empirical findings in those studies again.
A third research question relates to the effect of corporate profitability or free cash flow
and the degree of reliance on interpretive practices. (This is related to the issue of how to
fund or account for hermeneutic interpretive practices, briefly discussed earlier.) This
question arises as a result of the open-ended nature of the more hermeneutic interpretive
practices, and the parallel that can be drawn between these practices and advanced
scientific research. As we saw in the case of NDI's Color Studio, it is often difficult to
explain what the outcome of such activities will be. or even what a priori specific
questions are being addressed. In that sense, hermeneutic interpretive practices and
advanced scientific research are both activities that have a significant speculative aspect.
In the early 1990s, facing reduced profits due to the economic slowdown of the period,
and as a result of the reengineering movement that was at the height of its popularity
then, several highly regarded corporate research labs saw their funding reduced. More
252
importantly, the scope of their activities was drastically changed, in an effort to make
their work more directly relevant to the mainstream products and services of the parent
companies. Examples include the changes at General Motors Research Laboratory, IBM
Research, Bell Labs, and XEROX PARC.6 7 In a similar vein, are interpretive practices
viewed as a luxury in which companies indulge when they are flush with profits? The
case of NTC points to a connection between a cautious corporate stance and the
dominance of analytical practices. Although the cautious attitude in that case was blamed
on the fear of frittering an initial early success, it could as easily have been the result of
poor corporate performance.6 8
Finally, one of the more important questions that fell outside the scope of the present
study is the converse of the one raised in the previous paragraph. Namely, how does a
company's reliance on one or another of the different types of design approaches
presented in this dissertation ultimately affect corporate profitability? Are products
designed using a predominantly interpretive approach likely to be more or less successful
in the marketplace? What about critical success? And overall project profitability? These
"bottom line" questions are probably the ones of most interest to practitioners and
managers, but they are also the most difficult to answer. This is especially true in the case
of corporate profits, due to the countless confounding factors that may have little or
nothing to do with the design of the product, such as manufacturing overcapacity.
longstanding labor agreements. financial issues and so on.
Referring back to Clark and Fujimoto's work, one could use their metrics to assess the
effect of different practices on product development performance (Clark et al., 1991).
The metrics in question are development lead-time (or time to market), development
efficiency (the level of resources needed to develop a new product), and product integrity
(specifically, external integrity, i.e.. how closely product function and aesthetics meet the
67
More recently, this trend has seen a partial reversal, as corporate profits have risen in the late 1990s.
should be noted that when the empirical studies presented in this dissertation were conducted. Nissan,
the parent company of NDI, was already losing money ($1.3 billion in 1995 (Thornton et al.. 1997).) Yet,
among the three case studies presented in this dissertation, NDI is the site that relied most heavily on
hermeneutic interpretation (please refer to Chapter V.) This would seem to argue against the point of this
paragraph. That is not the case: it is clear from the case study that NDI has held a privileged position
within the Nissan Corporation from its inception, and that it has benefited from a hands-offattitude from
the parent company.
68 It
253
needs, wants, values, self-image, and lifestyle of the target customer.) One may very well
find that different practices or approaches to product development involve different
tradeoffs between development efficiency and product integrity. The problem with these
metrics however is that they work best when applied to development projects that can be
identified with a clear beginning and end, and a well-defined set of inputs and outputs.
Interpretive activities, by their nature, tend to transcend project boundaries. As we saw in
the case of NDI for example, and the Color Studio specifically, several of their activities
were not connected to particular projects but were ongoing processes that likely impacted
more than one project. Furthermore, an improved ability to interpret a situation and to
understand customer needs and wants are qualitatively different from a new
manufacturing process or a new technical feature. Unlike the latter, the costs and benefits
of the former are more difficult to measure and apportion among projects and product
lines. Developing the means for addressing these challenges is left for future research.
254
Bibliography
Adler, P. 1989. Technology Strategy: A Guide to the Literatures. In R. Rosenbloom, & R.
Burgelman (Eds.), Research on Technological Innovation, Management, and
Policy, Vol. 4: 25-151. Greenwich, CT: JAI Press.
Akin, O. 1978. How Do Architects Design? In G. Latombe (Ed.), Artificial Intelligence
and Recognition in Computer-AidedDesign: 27-34.New York: North Holland.
Akin, 0., & Lin, C. 1996. Design Protocol Data and Novel Design Decisions. In N.
Cross, H. Christiaans, & K. Dorst (Eds.), Analysing Design Activity: 35-63. New
York: John Wiley & Sons.
Akmajian, A., Demers, R. A., Farmer, A. K., & Harnish, R. M. 1995. Linguistics:
An
Introduction to Language and Communication (Fourth ed.). Cambridge: The
MIT Press.
Alexander, C. 1964. Notes on the Synthesis of Form. Cambridge, Massachusetts:
Harvard University Press.
Allen, T. J. 1971. Communications, Technology Transfer, and the Role of the Technical
Gatekeeper. R&D Management,
1: 14-21.
Allen, T. J. 1977. Managing the Flow of Technology. Cambridge, MA: The MIT Press.
Ancona, D. G., & Caldwell, D. F. 1990. Beyond Boundary Spanning: Managing External
Dependence in Product Development Teams. Journal of High Technology
Management Research, 1(119-135).
Ancona, D. G., & Caldwell, D. F. 1992. Bridging the Boundary: External Process and
Performance in Organizational Teams. Administrative Science Quarterly, 37:
634-665.
Armstrong, L. 1991. It Started with an Egg. Business Week(3242(Industrial/Technology
Edition)): 142,146.
255
Arnold, S. J., & Fischer, E. 1994. Hermeneutics and Consumer Research. Journal of
Consumer Research, 21: 55-70.
Audi, R. (Ed.). 1995. The Cambridge Dictionary of Philosophy. Cambridge: Cambridge
University Press.
Automotive News. 1995. Car guys obsolete, guru Drucker says. Automotive News,
February 7, 1995.
Autoweek. 1998. THE NEWS 8.24.98. AutoWeek, August 24, 1998, 4.
Axley, S. R. 1984. Managerial and Organizational Communication in Terms of the
Conduit Metaphor. Academy of Management Review, 9(3): 428-437.
Bailey, K. D. 1994. Typologiesand Taxonomies:An Introductionto Classification
Techniques. Beverly Hills, CA: Sage Publications.
Barron, K. 1998. Is it the sunshine? Forbes, September 21. 1998. 123-128.
Bayley, S. 1990. Harley Earl: Taplinger Publishing Co., Inc.
Bennet, J. 1994. The Designers Who Saved Chrysler. The New York Times, January 30,
1994.
Bennet, J. 1995. Four-Wheeling No Longer Just a Fad. The New York Times. February 9.
1995.
Beyer, J. M. 1992. Metaphors, Misunderstandings, and Mischief: A Commentary.
OrganizationScience, 3(4).
Boden, M. A. (Ed.). 1994. Dimensions of Creativity. Cambridge, MA: The MIT Press.
Bohn, J. 1993. Chrysler on the prowl. Automotive News, Aug 2, 1993, 1.
Bowen, K. H., Clark, K. B., & al., e. 1994. Development Projects: The Engine of
Renewal. Harvard Business Review, September-October 1994.
Brown, S. L., & Eisenhardt, K. M. 1995. Product Development: Past Research, Present
Findings, and Future Directions. Academy of Management Review, 20(2): 343378.
256
Bryant, A. 1992. Breaking the Mold at Nissan Design. New York Times Current Events
Edition, May 11, 1992, DI.
Bucciarelli, L. L. 1994. Designing Engineers. Cambridge, MA: The MIT Press.
Burchill, G., & Fine, C. H. 1997. Time versus market orientation in product concept
development: Empirically-based theory generation. Management Science, 43(4):
465-478.
Burchill,G. W. 1993.Concept Engineering:An Investigationof Time v. Market
Orientationin Product ConceptDevelopment.
Butler, R. 1997. Stories and Experiments in Social Inquiry. Organization Studies, 18(6):
927-948.
Carlile, P. R. 1997. Understanding Knowledge Transformation in Product
Development:Making KnowledgeManifest Through Boundary Objects.,
University of Michigan, Ann Arbor.
Carlile, P. R. 1999. Personal Communication.
Child. C. 1996a. Cherry: GM design plan is driven by customers. Automotive News,
September 23, 1996, 3.
Child, C. 1996b. GM plays it too safe, former design chief says. Automotive News.
September 23. 1996. 31.
Chomsky. N. 1957. Syntactic Structures. The Hague: Mouton.
Clark, K. B., & Fujimoto, T. 1990. The Power of Product Integrity. Harvard Business
Review, November-December
1990, 107-118.
Clark, K. B., & Fujimoto, T. 1991. Product Development Performance: Strategy,
Organization,and Managementin the WorldAuto Industry. Boston: Harvard
Business School Press.
Clausing,D. 1993. Total Quality Development:The DevelopmentofCompetitive New
Products (Draft dated August 22, 1993 ed.). Cambridge.
257
Clements, M. 1992. Altima Races into '93 Model Sales Lead. USA Today, Dec. 21, 1992,
1.
Connelly, M. 1990. Dodge Viper Gets Green Light. Automotive News, May 21, 1990.
Connelly, M. 1994. Clinics Pan LSE Rear, so Cadillac Revises it. Automotive News,
February 28, 1994, 1.
Cooper, R. G. 1990. Stage-Gate Systems: A New Tool for Managing New Products.
Business Horizons, May-June 1990, 44-54.
Cooper, R. G. 1994. Third-Generation
New Product Processes. Journal of Product
Innovation Management. 11(1): 3-14.
Cooper, R. G., & Kleinschmidt, E. J. 1986. An investigation into the new product
process. Journal of Product Innovation Management, 3: 71-85.
Cooper, R. G., & Kleinschmidt. E. J. 1987. New Products: What Separates Winners from
Losers? Journal of Product Innovation Management, 4: 169-184.
Coyne, R. 1995. Designing Information Technology in the Postmodern Age: From
Method to Metaphor. Cambridge: The MIT Press.
Coyne, R. 1998. Cyberspace and Heidegger's pragmatics. Information Technology &
People, 11(4): 338-350.
Coyne, R., & Snodgrass, A. 1991. Is designing mysterious? Challenging the dual
knowledge thesis. Design Studies, 12(3): 124-131.
Coyne, R., & Snodgrass, A. 1993a. Cooperation and Individualism in Design. Planning
and Design, 20: 163-174.
Coyne, R. D., Newton, S., & Sudweeks, F. 1993b. A Connectionist View of Creative
Design Reasoning. In J. S. Gero, & M. L. Maher (Eds.), Modeling Creativity and
Knowledge-Based Creative Design: 177-209. Hillsdale, NJ: Lawrence Erlbaum.
Coyne, R. D., Rosenman, M. A., Radford, A. D., Balachandran, M., & Gero, J. S. (Eds.).
1990. Knowledge-Based Design Systems. Reading, MA: Addison-Wesley.
258
Csikszentmihalyi,M. 1996.Creativity: Flow and the Psychologyof Discoveryand
Invention. New York, NY: Harper Collins Publishers.
Dascal, M. 1989. Hermeneutic Interpretation and Pragmatic Interpretation. Philosophy
and Rhetoric, 22(4).
Deschamps, J.-P., & Nayak, P. R. 1995. Product Juggernauts: How Companies
Mobilize to Generatea Stream of Market Winners. Boston,MA: Harvard
Business School Press.
Donaldson, L. 1992. The Weick Stuff: Managing Beyond Games. Organization Science,
3(4).
Dong, Q. 1999. Representing Information Flow and Knowledge Management in
ProductDesign Usingthe Design Structure Matrix. UnpublishedS.M.,
Massachusetts Institute of Technology, Cambridge.
Dorst, K., & Dijkhuis, J. 1996. Comparing Paradigms for Describing Design Activity. In
N. Cross, H. Christiaans, & K. Dorst (Eds.), Analysing Design Activity: 253-270.
Chichester: John Wiley & Sons.
Dougherty, D. 1990. Understanding New Markets for New Products. Strategic
ManagementJournal, 11:59-78.
Dougherty, D. 1992a. Interpretive Barriers to New Product Innovation in Large Firms.
Organization Science, 3(2): 179-202.
Dougherty, D. 1992b. A Practice-Centered Model of Organizational Renewal Through
Product Innovation. Strategic Management Journal, 13: 77-92.
Dougherty, D., & Heller, T. 1994. The Illegitimacy of Successful Product Innovation in
Established Firms. Organization Science. 5(2): 200-218.
Dreyfus,H. L. 1991.Being-in-the-World:A Commentaryon Heidegger'sBeing and
Time, Division I. Cambridge: The MIT Press.
Dreyfus, H. L. 1998. Why we do not have to worry about speaking the language of the
computer. Information Technology & People, 11(4): 281-289.
259
Eisenhardt, K. M. 1989. Building Theories From Case-Study Research. Academy of
Management Review, 14(4): 532-550.
Eppinger, S. D. 1991. Model-Based Approaches to Managing Concurrent Engineering.
Journal of Engineering Design, 2(4): 283-290.
Eppinger, S. D., Whitney, D. E., Smith, R. P., & Gebala, D. A. 1990. Organizing the
Tasks in Complex Design Projects. Paper presented at the ASME 2nd
International Conference on Design Theory and Methodology--DTM '90,
Chicago, Illinois.
Eppinger, S. D., Whitney, D. E., Smith, R. P., & Gebala, D. A. 1994. A Model-Based
Method for Organizing Tasks in Product Development. Research in Engineering
Design, 6: 1-13.
Eppinger. S. D., Whitney, D. E., & Yassine, A. A. 1999. The MIT Design Structure
Matrix Web Site, Vol. 2000.
Ericsson, K. A., & Simon, H. A. 1993. ProtocolAnalysis.
Cambridge, MA: The MIT
Press.
Fine, C. H., & Whitney,D. E. 1996.Is the Make-Buy DecisionProcessa Core
Competence?, MIT Center for Technology, Policy and Industrial Development.
Cambridge, MA.
Finke, R. A., Ward, T. B., & Smith, S. M. 1992. Creative Cognition: Theory, Research,
andApplications.
Cambridge, MA: The MIT Press.
Fischer, G., Nakakoji, K., Ostwald, J., Stahl, G., & Sumner, T. 1993. Embedding
Computer-BasedCriticsin the Contexts of Design. Paper presentedat the
InterCHI '93 Conference on Human Factors in Computing Systems, Amsterdam.
Fleming, A. 1991. Forward Thinking. Automotive News, Dec. 23, 1991, 4i.
Flint, J. 1992. Platform Madness. Forbes, Jan 20, 1992, 40.
Flint, J. 1996. Knees Together and Swing to the Right. Forbes, May 20, 1996, 3126.
260
Flint, J. 1997. Company of the year: Chrysler has the hot cars. More important, it has a
smart, disciplined management team. Forbes, Jan 13, 1997, 82.
French, M. J. 1985. Conceptual Designfor
Engineers. London: The Design Council.
Funes, P., & Pollack, J. 1997. Computer Evolution of Buildable Objects. Paper
presented at the Fourth European Conference on Aritficial Life.
Gadamer, H.-G. 1976. Philosophical Hermeneutics (D. E. Linge, Trans.). Berkeley:
University of California Press.
Gadamer, H.-G. 1981. Reason in the Age of Science (F. G. Lawrence, Trans.).
Cambridge: MIT Press.
Gallemore, T. W. 1. 1998. Virtual Workplace: When Metaphors Breakdown.
Unpublished M.S. Thesis, Massachusetts Institute of Technology, Cambridge,
MA.
Gardner,H. E. 1994.CreatingMinds : An Anatomy of CreativitySeen Throughthe
Lives of Freud, Einstein, Picasso,Stravinsky, Eliot, Graham, and Gandhi:
Basic Books.
Geertz, C. 1973. The Interpretation of Cultures: Selected Essays. New York: Basic
Books.
Gelsi, S. 1996. Robert Lutz. Mediaweek, Oct 7, 1996, 104.
Gero, J. S. (Ed.). 1989. Artificial Intelligence in Design. Berlin: Springer-Verlag.
Gero, J. S., & Kazakov, V. 1996. Evolving Building Blocks for Design Using Genetic
Engineering: A Formal Approach. In J. S. Gero (Ed.), Advances in Formal
Design Methods for CAD: 31-50. London: Chapman and Hall.
Gioia, D. A., & Pitre, E. 1990. Multiparadigm Perspectives on Theory Building.
Academy of Management Review, 15(4): 584-602.
Giv6n,T. 1989.Mind, Code,and Context:Essays in Pragmatics.Hillsdale:Lawrence
Erlbaum Associates, Inc., Publishers.
261
Griffin, A., & Hauser, J. R. 1990. Patterns of Communication among Marketing,
Engineering and Manufacturing -- A Comparison Between Two New Product
Teams.MIT InternationalCenterfor Research on the Managementof
Technology WP#16-90.
Hauser, J. R., & Clausing, D. 1988. The House of Quality. The Harvard Business
Review, March-June 1988, 63-73.
Hesse, M. 1980.Revolutions and Reconstructionsin the Philosophy of Science.
Bloomington, IN: Indiana University Press.
Hirschman, E. C. 1986. Humanistic Inquiry in Marketing Research: Philosophy, Method,
and Criteria. Journal of Marketing Research, 23: 237-249.
Hirshberg,J. 1998.The CreativePriority:Driving InnovativeBusiness in the Real
World. New York: Harper Business.
Holbrook, M. B., & Grayson, M. W. 1986. The Semiology of Cinematic Consumption:
Symbolic Consumer Behavior in Out of Africa. Journal of Consumer Research,
13: 374-381.
Hollins, B., & Pugh, S. 1990. Successful Product Design: What to Do and When.
London: Butterworths.
Hubka, V., & Eder, W. E. 1988. Theory of Technical Systems: A Total Concept Theory
for Engineering Design. Berlin:Springer-Verlag.
Hudson, L. A., & Ozanne, J. L. 1988. Alternative Ways of Seeking Knowldge in
Consumer Research. Journal of Consumer Research, 14: 508-521.
Jantsch, E. 1980.The Self-OrganizingUniverse:Scientific and Human Implicationsof
the Emerging Paradigmof Evolution. Oxford:PergamonPress.
Jonas, W. 1993. Design as problem-solving? or: Here is the solution -- what was the
problem? Design Studies, 14(2).
Jones, J. C. 1981. Design Methods: Seeds of Human Futures (Second ed.). New York:
John Wiley & Sons.
262
Katz, R. 1982. The Effects of Group Longevity on Project Communication and
Performance. Administrative Science Quarterly, 27: 81-104.
Katz, R., & Tushman, M. L. 1981. An Investigation into the Managerial Roles and Career
Paths of Gatekeepers and Project Supervisors in a Major R&D Facility. R&D
Management,
11: 103-110.
Keebler, J. 1993. Chrysler moves forward with Viper coupe, Prowler studies. Automotive
News, Jul 5, 1993, 3.
Kim, D. H. 1993.A Framework and Methodologyfor Linking Individual and
Organizational Learning: Applications in TQM and Product Development.
Unpublished Ph.D., Massachusetts Institute of Technology, Cambridge.
Knight. F. H. 1957. Risk, Uncertainty, and Profit. New York: Kelley and Millman.
Knoll, B. 2001. More Sport (and Less Utility) Than the Basic S.U.V. The New York
Times, March 11, 2001.
Krippendorff, K. 1989. On the Essential Contexts of Artifacts or on the Proposition that
"Design is Making Sense (of Things)". Design Issues, 5(2): 9-39.
Krippendorff, K., & Butter, R. 1989. Product Semantics: Exploring the Symbolic
Qualities of Form. IDSA.
Krueger, R. A. 1994. Focus Groups: A Practical Guidefor Applied Research. Thousand
Oaks, California: Sage Publications.
Kuhn, T. 1970. The Structure of Scientific Revolutions. Chicago: University of Chicago
Press.
Lakoff, G. 1987. Women, Fire, and Dangerous Things: What Categories Reveal about
the Mind. Chicago: The University of Chicago Press.
Lakoff, G., & Johnson. M. 1980. Metaphors We Live By. Chicago: The University of
Chicago Press.
Leonard, D., & Rayport, J. F. 1997. Spark innovation through emphatic design. Harvard
Business Review, November-December
263
1997, 102-113.
Leonard, D., & Straus, S. 1997. Putting your company's whole brain to work. Harvard
Business Review, July-August 1997, 11 1-120.
Lester, R. K., Piore. M. J., & Malek, K. M. 1998. Interpretive Management: What
General Managers can Learn from Design. Harvard Business Review, MarchApril 1998.
Lindinger, H. (Ed.). 1990. Ulm Design: The Morality of Objects (Hochschule fir
Gestaltung UIm 1953-1968). Cambridge, MA: The MIT Press.
Lutz, R. A. 1998. Guts: The Seven Laws of Business That Made Chryslerthe World's
HlottestCar Company: John Wiley & Sons.
MacKenzie, A. 1993. East meets west on the drawing board. Australian Business
Monthly, 13(9): 128-131.
Malek. K. M. 1996.A Case Study in Product Development: The 1984-199x Chevrolet
Corvette. Working Paper No. MIT IPC 96-003WP, MIT Industrial Performance
Center, Cambridge, MA.
Marquis, D. G. 1982. The Anatomy of Successful Innovation. In M. L. Tushman. & W.
L. Moore(Eds.),Readings in the Management of Innovation:42-50.
Cambridge, MA: Ballinger.
Maskery, M. A. 1996. Heavy Halo. Automotive News. Oct. 21, 1996. 29i.
McClelland, J. L., & Rumelhart, D. E. 1988. Explorations in Parallel Distributed
Processing:ComputationalModels of Cognitionand Perception.Cambridge.
MA: The MIT Press.
McCord,K., & Eppinger,S. D. 1993.Managing the Integration Problemin Concurrent
Engineering.
No. Working Paper no. 3594. MIT Sloan School of Management,
Cambridge, Massachusetts.
Mey, J. L. 1993. Pragmatics: An Introduction. Oxford: Blackwell Publishers.
Meyer, A. D. 1984. Mingling Decision Making Metaphors. Academy of Management
Review,
9(1): 6-1 7.
264
Minsky, M. L. 1988. The Society of Mind. New York: Simon & Schuster.
Mintzberg, H. 1979. An Emerging Strategy of "Direct" Research. Administrative Science
Quarterly, 24(4): 582-589.
Mitchell, W. 1996. Personal Communication.
Mitchell, W. J. 1990. The Logic of Architecture:
Design, Computationand
Computation. Cambridge: The MIT Press.
Moenaert, R. K., & Souder, W. E. 1990. An Information Transfer Model for Integrating
Marketing and R&D Personnel in New Product Development Projects. Journal of
Product Innovation Management, 7: 91-107.
Morgan, G., & Smircich, L. 1980. The Case for Qualitative Research. Academy of
Management Review, 5: 491-500.
Myers, s., & Marquis, D. G. 1969. successful Industrial Innovations. No. NSF 69-17,
National Science Foundation, Washington, DC.
Navinchandra,D. 1991.Exploration and Innovation in Design: Towardsa
Computational Model. New York: Springer-Verlag.
Nussbaum, B. 1993. Hot Products: Smart Design is the Common Thread. Business
Week, June 7, 1993, 54-57.
Nussbaum, B. 1995. Winners: The Best Product Designs of the Year. Business Week,
June 5, 1995, 88-91.
Nussbaum, B. 1997. Winners: The Best Product Designs of the Year. Business Week,
June 12, 1997, 92-95.
O'Connor, P. 1994. Implementing a Stage-Gate Process: A Multi-Company Perspective.
Journal of Product Innovation Management, 11(3):183-200.
Pahl, G., & Beitz, W. 1984. Engineering Design (K. Wallace, Trans.). London, Berlin:
The Design Council, Springer-Verlag.
265
Palmer, R. E. 1969. Hermeneutics: Interpretation Theory in Schleirmacher, Dilthey,
Heidegger, and Gadamer. Evanston: Northwestern University Press.
Passell, P. 1998. An Outback Gone Upscale. The New York Times, February 7, 1998.
Phillips, D. 1998. Competition Gets Intense on Minivans. The Detroit News, April 28,
1998, April 28, 1998.
Pierce, J. R. 1980.An Introduction to Information Theory: Symbols, Signals and Noise
(Second ed.). New York: Dover Publications.
Pimmler, T. U., & Eppinger, S. D. 1994. Integration Analysis of Product
Decompositions. Paper presented at the 6th International Conference on Design
Theory and Methodology - DTM '94.
Piore, M. J. 1979. Qualitative Research Techniques in Economics. Administrative
Science Quarterly, 24(4): 560-569.
Piore, M. J., Lester, R. K., Kofman, F. M.. & Malek, K. M. 1994. The Organization of
Product Development. Industrial and Corporate Change, 3(2): 405-434.
Piore, M. J., Lester, R. K., & Malek, K. M. 1995. Case Studies in the Organization of
ProductDevelopment:The Cellular TelephoneEquipment Industry. Working
Paper No. MIT IPC 95-008WP, Massachusetts Institute of Technology, Industrial
Performance Center, Cambridge.
Piore, M. J., Lester, R. K., & Malek, K. M. 1997a. The Division of Labor, Coordination,
and Integration: Case Studies in the Organizationof Product Design in the
Blue Jeans Industry. Working Paper No. MIT IPC 97-005WP, Massachusetts
Institute of Technology, Industrial Performance Center, Cambridge.
Piore, M. J., Lester, R. K., & Malek, K. M. 1997b. The Division of Labor, Coordination,
and Integration.CaseStudies in the Organizationof Product Development:
Medical Instruments. Working Paper No. MIT IPC 97-007WP, Massachusetts
Institute of Technology, Industrial Performance Center, Cambridge.
266
Prahalad, C. K., & Hamel, G. 1990. The Core Competence of the Corporation. Harvard
Business Review, May 1990.
Pugh,S. 1991. TotalDesign: IntegratedMethodsfor Successful ProductEngineering.
Wokingham: Addison-Wesley Publishing Company.
Random House. 1992. Random House Webster's Electronic Dictionary and Thesaurus,
College Edition, 1.0 ed.: Reference Software International.
Rechtin, M. 1999. Nissan's chief designer to star in new ads. Advertising Age, 70(16): 33.
Rechtin, M., Crate, J. R., & Jewett, D. 1997. Nissan answers Toyota, slashes up to $1,500
from Altima sticker. Automotive News, June 23, 1997, 4-5.
Reddy, M. J. 1979. The Conduit Metaphor. In A. Ortony (Ed.), Metaphor and Thought:
284-324. Cambridge: Cambridge University Press.
Robertson. D., & Ulrich, K. 1998. Planning for Product Platforms. Sloan Management
Review, 39(4): 19-31.
Rosch, E., & Lloyd, B. B. 1978. Cognition and Categorization. Hillsdale, N.J.: L.
Erlbaum Associates.
Rosenberg, R. 1997. By design, these firms take on other companies' products. The
Boston Globe, May 11, 1997.
Rothwell,R. 1972.Factorsfor Success in Industrial Innovationsfrom Project
SAPPHO- A Comparative
Studyof Successand Failurein Industrial
Innovation., S.P.R.U., Brighton, Sussex, England.
Rothwell, R., Freeman, C., Horsley, A., Jervis, V. T. P., Robertson, A., & Townsend, J.
1974. SAPPHO Updated -- Project Sappho Phase II. Research Policy, 3: 258-291.
Rowe, P. G. 1987. Design Thinking. Cambridge, MA: The MIT Press.
Rubin, H. J., & Rubin, 1. S. 1995. Qualitative Interviewing: The Art of Hearing Data.
Thousand Oaks, CA: Sage Publications.
Sch6n, D. A. 1983. The Reflective Practitioner. New York: Harper Collins.
267
Searle, J. R. 1985. Expression and Meaning: Studies in the Theory of Speech Acts.
Cambridge (Eng.): Cambridge University Press.
Searle, J. R. 1992. The Rediscovery of the Mind. Cambridge: The MIT Press.
Seely Brown, J., & Duguid, P. 1991. Organizational Learning and Communities-ofPractice: Toward a Unified View of Working, Learning, and Innovation.
Organization Science, 2(1): 40-57.
Shannon, C. E., & Weaver, W. 1949. The Mathematical Theory of Communications.
Urbana: University of Illinois Press.
Siano, J. 1998. Daewoo Who? Some Low-Key Answers. The New York Times, July 19,
1998.
Simon, H. A. 1981. The Sciences of the Artificial (Second ed.). Cambridge, MA: The
MIT Press.
Snodgrass, A., & Coyne, R. 1990. Is Designing Hermeneutical. Working Paper.
Departement of Architecture. University of Sidney, Sidney.
Snodgrass, A., & Coyne, R. 1992. Models. Metaphors and the Hermeneutics of
Designing. Design Issues, IX( 1): 56-74.
Stahl, G. 1993. Interpretation in Design: The Problem of Tacit and Explicit
Understanding in Computer Support of Cooperative Design. Unpublished Ph.D.
Thesis, University of Colorado.
Stalk, G., Evans, P., & Shulman, L. E. 1992. Competing on Capabilities: The New Rules
of Corporate Strategy. The Harvard Business Review, March-April 1992.
Staudenmayer. N. A. 1997. Managing M.ultipleInterdependencies in Large Scale
Software Development Projects. Unpublished Ph.D., Massachusettes Institute of
Technology, Cambridge, MA.
Steward, D. V. 1981. The Design Structure System: A Method for Managing the Design
of Complex Systems. IEEE Transactions on Engineering Management, EM28(3): 71-74.
268
Stiny, G. 1980. Introduction to Shape and Shape Grammars. Environment
and Planning
B, 7: 343-351.
Stiny, G., & Mitchell, W. J. 1978. The Palladian Grammar. Environment and Planning,
B 5(1): 5-18.
Stout, R. H. 1995. Car guys obsolete? Drucker's off base. Automotive News, April 10,
1995.
Strauss, A. L. 1987. Qualitative AnIalysisfor Social Scientists. Cambridge; New York:
Cambridge University Press.
Suh, N. P. 1990. The Principles of Design. New York: Oxford University Press.
Tagliabue, J. 2001. Italian by Design; Stylists of Turin Expand Impact of Auto
Couturiers. The New York Times, January 10, 2001.
Takeuchi, H., & Nonaka, 1. 1986. The New New Product Devleopment Game. Harvard
Business Review, 64(1): 137-146.
Taras, J. J. 1994. What's the Big Deal About 'Cab Forward' Cars? New York Times, Dec.
I1, 1994, 111.
Taylor, A. I. 1993. Hitmaker from the Parts Department. Fortune, Mar. 22, 1993. 177.
Taylor, A. I. 1994. Will Success Spoil Chrysler. Fortune, Jan. 10, 1994, 88.
Thompson, C. J., Locander, W. B., & Pollio, H. R. 1989. Putting Consumer Experience
Back into Consumer Research: The Philosophy of Existential Phenomenology.
Journal of Consumer Research, 16.
Thompson, C. J., Pollio, H. R., & Locander, W. B. 1994. The Spoken and the Unspoken:
A Hermeneutic Approach to Understanding the Cultural Viewpoints that Underlie
Consumer's Expressed Meanings. Journal of Consumer Research, 21.
Thornton, E., & Armstrong, L. 1997. Nissan's Slow U-Turn; Its recovery is far from
complete. Business Week, May 12, 1997, 54.
269
Tong, C., & Sriram, D. (Eds.). 1992a. Artificial Intelligence in Engineering Design:
Design Representationand Models of Routine Design.(Vol. 1). Boston:
Academic Press.
Tong, C., & Sriram, D. (Eds.). 1992b.Artificial Intelligence in Engineering Design:
Models of Innovative Design,Reasoning about PhysicalSystems, and
Reasoning about Geometry. (Vol. 2). Boston: Academic Press.
Tovey, M. 1992. Intuitive and Objective Processes in Automotive Design. Design
Studies, 13(1): 23-41.
Ullman, D. G. 1992. The Mechanical Design Process. New York: McGraw-Hill
Inc.
Ulrich, K. T., & Eppinger, S. D. 1995. Product Design and Development. New York:
McGraw-Hill, Inc.
Van Maanen, J. 1988. Tales of the Field: On Writing Ethnography. Chicago: The
University of Chicago Press.
Vasilash. G. S. 1991. Teams Are Workin& Hard at Chrysler. Production. Oct 1991, 50.
Wallace. D. R. 1991a. A Computer Model ofAesthetic
Industrial Design. Unpublished
Master of Science, Massachusetts Institute of Technology, Cambridge,
Massachusetts.
Wallace, D. R. 1991b. Personal Communication.
Washington, F. 1995. Design 2000: bold, sharp-edged lines are coming. Ward's Auto
World, 31(11): 31.
Weick, K. E. 1995. Sensemaking
in Organizations. Thousand Oaks. CA: Sage
Publications.
Wheelwright, S. C., & Clark, K. B. 1992. Revolutionizing
Product Development:
Quantum Leaps in Speed, Efficiency, and Quality. New York: The Free Press.
Woodruff, D. 1995. A Van For All Seasons. Business Week. June 5, 1995, 97.
270
WSJ. 1999. What's Black and White and Read All Over and Redesigned by Nissan? The
Wall Street Journal, July 8, 1999.
Yin, R. K. 1994. Case Study Research: Design and Methods (Second ed.): Sage
Publications.
Yung, K. 1997. Chrysler: You gotta have heart. No.3 automaker won't join brand
manager movement. Detroit News, June 29, 1997.
Zirger, B. J., & Maidique, M. 1990. A Model of New Product Development: An
Empirical Test. Management Science, 36(867-883).
271
THESIS PROCESSING SLIP
FIXED FIELD:
ill.
name
biblio
index
-
· COPIES:
-
ArchiveW'
Aero
Dewey
'arker
Lindgren
Music
Rotch
Science
TITLE VARIES:
E
NAME VARIES:
"COi
IMPRINT:
Hum
Sche-Plough
-
I
,
'.G.
L/gws
2
(COPYRIGHT)_
Do,COLLATION:
ML AT.nN
b.
. ADD: DEGRE
b-ADD:
DEGREE:
*· DEPT.:
b,
·o SUPADD
b-ADD:
DEGRE
DEGREE:
:
DET.:
SUPERVISOF
SUPERVISORS:
-
10,
b,NIEOTES:
NOTES:
Ik
Ik
A .,
/'
cat'r:
I
loDEPT:· , DEPT:
DEPT:
loYEAR:
EAR:
· YEAR:
I,,
I,,
10,NAE
· NAME:
*'
NAME: -b
-
--
"I/I
--110
date:
page:
-
i1I
i-1
P'f i
/
L/ 'f
,,DEGREE:
I-DEGREE:
DJO/
/
6--1,
,
~'.
_